WEBVTT 00:24:00.000 --> 00:00:40.330 I want to first start by thanking our stationary acoustic working group who really formulated the idea and vision for this workshop and umm and put it all together in our hosting it for us today. 00:00:40.340 --> 00:01:00.600 So thank you to all of them for all the labor that has gone into that and the Intel that they're sharing with everybody, a little bit of background on the inspiration for this workshop and where it kind of originated was the as we were all aware, there are a lot of challenges to managing acoustic data and the more you have, the harder it gets. 00:01:01.390 --> 00:01:11.020 But there are a lot of folks who have worked to figure out streamlined processes, tools, tips and tricks that are helping things work at least a little bit better. 00:01:11.030 --> 00:01:14.330 Or a little bit smoother for them, or in some instances a lot. 00:01:14.570 --> 00:01:21.720 And so we just thought we would start with, let's share that information more broadly, insert building connections. 00:01:22.150 --> 00:01:29.760 Uh, it within the community on kind of what's out there and what are some options that folks perhaps could adopt and how could we grow from there? 00:01:30.410 --> 00:01:32.060 So that's where we're starting today. 00:01:32.810 --> 00:01:34.600 We do have a really tight agenda. 00:01:34.610 --> 00:01:38.700 There's a lot of content to get through and and for awareness. 00:01:38.710 --> 00:01:41.530 We aren't going to be doing much of a deep dive on any of it. 00:01:41.540 --> 00:01:43.680 You'll get a taste on all of these topics. 00:01:44.610 --> 00:01:45.090 Umm. 00:01:45.430 --> 00:02:02.340 And then we're gonna be looking for information or input from you at the end, which we also did on Tuesday on what do you see as good next steps for how to start putting this into practice or getting these tools or resources into additional folks hands? 00:02:03.270 --> 00:02:15.440 Umm, the also to note that due to the very tight agenda, Andrea is gonna work to keep us all on time so that we don't run short and we get through everything on the agenda. 00:02:15.450 --> 00:02:21.850 So she may cut off a presenter or speaker and may abbreviate our Q&A portions as needed. 00:02:23.060 --> 00:02:30.590 Uh to help facilitate responding to questions that you may have, we encourage you to put any questions that you have into the chat through the meeting. 00:02:30.600 --> 00:02:38.130 And as folks are able, they'll respond directly there and then we'll get to as many as we can in the Q&A portion. 00:02:38.140 --> 00:02:44.000 NABat we aren't able to address, we will work to respond to in writing after the fact. 00:02:44.810 --> 00:02:45.520 Umm. 00:02:45.570 --> 00:02:46.760 And I think that's it. 00:02:46.770 --> 00:02:50.350 So with that, I'll pass things over to our first presenter. 00:02:53.190 --> 00:02:53.600 Great. 00:02:53.610 --> 00:02:54.460 Thanks, Bethany. 00:02:54.470 --> 00:02:58.610 So Donald, you're up to give us strategies for manual vetting. 00:03:05.460 --> 00:03:05.980 Can see it. 00:03:05.990 --> 00:03:06.760 Donald looks great. 00:03:08.430 --> 00:03:10.210 Uh, turn your mic on now. 00:03:18.320 --> 00:03:18.770 That'll help. 00:03:20.360 --> 00:03:20.860 OK. 00:03:20.440 --> 00:03:21.010 OK. 00:03:21.340 --> 00:03:24.240 Second try, that's always something. 00:03:22.560 --> 00:03:25.450 See, there we go. OK. 00:03:24.250 --> 00:03:25.830 OK. Who? 00:03:26.160 --> 00:03:26.650 Alright. 00:03:26.860 --> 00:03:27.590 So, yeah. 00:03:27.660 --> 00:03:29.590 Hey everybody, I'm Donald solek. 00:03:29.600 --> 00:03:41.230 I'm the owner of Vesper bat detection services, where a consulting compNABat provides training on manual vetting, or we'll take your data and manually vet it ourselves. 00:03:41.240 --> 00:03:50.780 If you don't have the time or expertise or interest in doing it yourself, so I'm going to start off by talking about uh, the workflow of manual bedding. 00:03:51.430 --> 00:03:52.670 But first, I'd like to just. 00:03:54.790 --> 00:03:57.570 Kind of emphasize why we do this to ourselves. 00:03:59.140 --> 00:03:59.650 Wait, what? 00:04:00.060 --> 00:04:10.250 What the importance of this is so a few years ago I tested the accuracy of of auto classifiers using over 1000 reference calls by 9 Northeastern Pass species. 00:04:10.700 --> 00:04:15.490 I presented this at the Wildlife Society meeting and I'm currently writing a manuscript on this. 00:04:16.370 --> 00:04:16.840 Umm. 00:04:17.470 --> 00:04:27.330 And there's a lot going on here, but the gist is we found that sauNABat only performed well for all 5 metrics of accuracy that we looked at for just five of the nine species. 00:04:28.290 --> 00:04:44.310 Glascoe Pro only perform well for four of the species, and bcid didn't perform well for any species, and these are relatively high quality hand-picked reference calls, so the deck was stacked in favor of these classifiers and they still did not perform exceptionally well. 00:04:44.500 --> 00:04:50.910 So the implication is that they would not perform, they perform even poorer on real world data, which tends to be messier. 00:04:52.700 --> 00:04:54.250 So take home messages. 00:04:54.310 --> 00:05:00.870 Auto classifiers can be helpful, but files really need to be manually vetted if you want to have confidence in the identifications. 00:05:02.050 --> 00:05:09.180 Unfortunately, despite the importance of manual verification of back calls, there are very few resources for learning the skill in North America. 00:05:09.650 --> 00:05:22.940 This image comes from one of the classes I teach showing diagnostic characteristics of four Western species, Janet Tipuric and Corey Lawson also teach classes, and sometimes Joe Susak and Chris Corbin hold workshops, but that's it. 00:05:23.130 --> 00:05:34.050 There's no official certification or permit required to do this work, and nobody's evaluated consistency among manual reviewers, so it's really the wild, Wild West out there in terms of manual vetting. 00:05:34.940 --> 00:05:38.910 So that said to hopefully make things a little more civilized. 00:05:39.720 --> 00:05:47.240 This would be my recommended workflow when sitting down and doing manual vetting yourself, and this comes from one of my classes. 00:05:48.500 --> 00:05:54.370 So when you're confronted with an unknown call for the first time, the first question you should ask yourself is geography. 00:05:54.680 --> 00:05:55.330 What? 00:05:55.380 --> 00:05:56.750 Where was this call collected? 00:05:56.760 --> 00:06:00.560 What are the possible species that could occur there that will narrow your list of suspects? 00:06:01.460 --> 00:06:03.810 Next, you should assess the quality of the call. 00:06:04.200 --> 00:06:05.890 Was the detector plays in good? 00:06:06.580 --> 00:06:08.390 Does the recording contain a single bat? 00:06:08.460 --> 00:06:10.970 Does it contain five or more clear search phase pulses? 00:06:10.980 --> 00:06:12.150 Does it have some harmonics? 00:06:12.620 --> 00:06:16.190 If not, you might not want to bother trying to identify that file. 00:06:16.280 --> 00:06:25.510 If it does, then you can go into the next step, which is to look for the minimum frequency of the calls, which will help narrow down your list of suspects even further. 00:06:25.980 --> 00:06:30.850 You can then evaluate the pattern of the minimum frequency across the sequence of calls. 00:06:30.860 --> 00:06:36.280 Is it steady or does it fluctuate and minimum frequency because that will also narrow down your playing field. 00:06:36.710 --> 00:06:43.650 You can evaluate shape, which sometimes provides good information on species ID and finally you reach the moment of truth. 00:06:45.700 --> 00:06:52.600 Are you confident enough to label this file as a species and defend it in court? 00:06:53.110 --> 00:07:02.780 Or do you want to be more conservative in place it in a in a species group, and then you just rinse, wash and repeat for the next file? 00:07:03.130 --> 00:07:05.770 So hopefully that was helpful. 00:07:07.120 --> 00:07:12.180 Are now going to jump into sohbat terms and workflow to kick off. 00:07:13.030 --> 00:07:17.680 Ohh, the digital portions of of of this section. 00:07:17.750 --> 00:07:21.600 So this is sahabat. 00:07:21.650 --> 00:07:24.940 When you open it for the first time, this is technically called the Sabbat viewer. 00:07:25.870 --> 00:07:28.380 And we're not going to be spending much time here. 00:07:28.430 --> 00:07:50.550 Instead, we're going to be looking at something called the the Sano vet tool, which you can get to by clicking the the button here in the lower left corner of the viewer, and you should use this on obet tool once you've already scrubbed your data for noise and run it through auto classifier when you open the table and import your data, you'll get something that looks like this. 00:07:50.560 --> 00:07:59.060 I've gotta just a truncated portion for for demonstration purposes, but this this this sono vet has five tabs. 00:07:59.560 --> 00:08:10.600 We're right now in the bedding table tab and the vetting table contains dozens of columns and most of them aren't gonna be that useful to you. 00:08:10.970 --> 00:08:20.580 So what I wanna do is highlight the 12 columns that will be useful for manual bedding, particularly for NABat starting with the species manual ID column. 00:08:21.110 --> 00:08:22.290 This ones pretty straightforward. 00:08:22.300 --> 00:08:23.960 This is where you when you're vetting. 00:08:23.970 --> 00:08:28.400 This is where you would put your label for any files you felt confident in. 00:08:28.650 --> 00:08:29.600 Right now it's blank. 00:08:29.690 --> 00:08:30.370 We'll get back to that. 00:08:31.910 --> 00:08:41.260 The next column over is the species except column and this is where Sahabat will label files that it is confident in. 00:08:41.590 --> 00:08:46.770 So this first file here it is, put a four four letter species code of anpa. 00:08:46.780 --> 00:08:47.520 That's a palette. 00:08:47.530 --> 00:08:48.380 That song? 00:08:48.390 --> 00:08:49.100 That's pretty confident. 00:08:49.110 --> 00:08:53.010 That's a palette that next one of things is a town's bigger bad. 00:08:53.550 --> 00:08:55.580 There's some blanks here too, and those are ones. 00:08:55.590 --> 00:08:58.560 Those are files that song that just wasn't confident about. 00:09:00.360 --> 00:09:14.110 Sonipat gets his confidence for these next two columns, so starting with this one on the on the right number of accepted pulses when you run your data through the auto classifier and sauNABat, you can choose how many pulses it examines to make a species ID. 00:09:14.400 --> 00:09:15.810 The default is 32. 00:09:15.900 --> 00:09:22.250 For this run I use 16 and so this number 13 here is telling us that somebody tried to look for 16 good pulses. 00:09:22.320 --> 00:09:28.350 It only found 13 and then this other number here, this eight in the number of majority pulses here. 00:09:28.430 --> 00:09:32.270 This eight tells us that eight of those came out of those thirteen came out as palette. 00:09:32.280 --> 00:09:33.900 Bat OK. 00:09:34.860 --> 00:09:35.910 Umm, the next. 00:09:36.060 --> 00:09:43.570 So the bigger back below that codo thought, it was very confident in eight of eight pulses came out as that species and the third one down. 00:09:43.580 --> 00:09:49.310 It was also confident in this being a bigger bat, although this was three of three, so some of that was really confident. 00:09:49.320 --> 00:09:52.910 But with only three pulses, that's one we should probably be skeptical about. 00:09:53.520 --> 00:09:55.440 But we're manually reviewing all these anyway. 00:09:55.640 --> 00:10:10.820 But just something to think about the next column over is the till day spa column and this is where song about will put its best guess for species that it wasn't confident in labeling in the species except column. 00:10:11.150 --> 00:10:13.840 So here it left this. 00:10:13.850 --> 00:10:20.880 You know, I left this up as a blank in in this column, but here it's saying that I'm leaning heavily towards this being an FU or a big brown bag. 00:10:21.040 --> 00:10:28.430 That's because five of the 10 pulses came out as this species, and so I find this column is different than what you get with glidescope. 00:10:28.440 --> 00:10:36.360 This is like a step above and and really I think elevates on a bat in terms of being able to do manual review. 00:10:38.340 --> 00:10:47.690 Next column over is what's called the first column, and this just combines the species labels from the species except and they till they spa column so that everything's just in one. 00:10:47.780 --> 00:10:57.750 Everything that song about thought was a bat and deserve a species label, either because it was confident or it was leaning towards it, goes here and this becomes more important when we get to NABat stuff. 00:10:59.050 --> 00:11:01.100 The next three columns second, third and 4th. 00:11:01.300 --> 00:11:11.600 These you can consider as alternate ID, so maybe this first one isn't a pallet bat but there were pulses in here that could have been silver haired Mexican, free tailed or hoary bat. 00:11:11.670 --> 00:11:14.230 So maybe that's the speeches you're looking at. 00:11:14.850 --> 00:11:18.640 This will also capture species files that had more than two bats in it. 00:11:18.930 --> 00:11:27.170 Maybe the first species will come out as a big brown because most of the pulses came out as big Brown, but there was a red bat flying in the background and it might capture that as well. 00:11:28.110 --> 00:11:32.330 What these columns are best for are for mining your data for rare species. 00:11:32.670 --> 00:11:35.290 Let's say you're really interested in little brown bats species. 00:11:35.300 --> 00:11:41.280 Code is MYLU and you didn't find any in species except column that till they saw the first column. 00:11:41.470 --> 00:11:53.560 But you do see there's at least one in the second column, so you could sort the second column alphabetically, go to the MMS and look to see how many to see which files might have been hiding a little brown bat in amongst there. 00:11:53.570 --> 00:11:55.780 So that's another great utility of sauna. 00:11:55.790 --> 00:11:58.910 Bat is being able to mine your data for rare species. 00:12:00.870 --> 00:12:05.040 We're still in the betting table now, shifted over to the right to look at three more columns. 00:12:05.120 --> 00:12:07.680 The first two are apparent dirt and next DIR up. 00:12:08.170 --> 00:12:10.600 These are just the directories where your data lives. 00:12:10.850 --> 00:12:22.420 For NABat data, parent dura corresponds with the quadrant that you collected data in and next door up is the grid cell and then monitoring night just indicates which night the data were collected on. 00:12:23.290 --> 00:12:27.000 So now we have a better understanding of these 12 columns. 00:12:27.370 --> 00:12:35.230 But now I want to talk about our workflow and how we can reorganize these data to make our labeling and vetting very efficient. 00:12:35.460 --> 00:12:47.220 So to do this, we go over to the Settings tab and in the setting tab we have this Gray bar here on the left which is where you could set up your own sort template to sort things however you want to. 00:12:47.930 --> 00:12:50.940 I'm going to provide the template. 00:12:50.950 --> 00:12:53.900 I use this is something I got from the Northwest bathtub and it is gold. 00:12:53.950 --> 00:12:59.500 I highly recommend taking you screenshot right now and using this whenever you do NABat in Sano vet. 00:13:00.110 --> 00:13:10.350 Remember that NABat they require, they want voucher calls for each species by each detector by each night, and so this is sorting your data by species detector and knife. 00:13:11.270 --> 00:13:35.040 These two rows here the pulses roads that's thrown in so that you can organize your data by least likely or most confident to least confident and so this can really improve your workflow if you got a lot of data and a lot of it came out as Mexican freetail bats, you can look at all your freetail bats by different detectors and different nights in the order of which is most likely to least likely. 00:13:35.110 --> 00:13:44.590 So you can just look at the first one, say, yeah, that's actually retail back make that your browser call and then skip to the next night and that can save tons of time looking through data that you don't need to bother with. 00:13:45.620 --> 00:13:49.730 So we set up our template, we click sort, we go back 01. 00:13:49.740 --> 00:13:58.390 Last thing, sorry if you use this, be sure to sort these Z to a so you're sorting most likely to least likely rather than least likely to most likely. 00:13:58.580 --> 00:14:17.290 So now we sorted go back to our table and now Sonipat has organize things by species here in the first column alphabetically and by directory or by quadrant, and by monitoring night and so this first file Sonipat leaves as a pallet bat this sauna. 00:14:17.300 --> 00:14:22.150 That table is connected is directly linked to the song about viewer, which is off screen right now. 00:14:22.160 --> 00:14:28.420 It's best to have that on a second screen so you can do this simultaneously, but if you click on this file it would show up in that screen. 00:14:28.430 --> 00:14:29.850 You could look at it and you go well. 00:14:29.860 --> 00:14:32.270 Actually, I think that's just a unknown low frequency. 00:14:32.280 --> 00:14:32.490 Bad. 00:14:33.240 --> 00:14:33.730 Umm. 00:14:34.000 --> 00:14:35.390 And then you move on to the next one. 00:14:35.440 --> 00:14:37.620 Which song about that very confidently was? 00:14:37.630 --> 00:14:44.190 It counts as bigger bat and we look at it, we agree and that was for the Northwest quadrant on May 13th. 00:14:44.260 --> 00:14:52.370 We see there's another confident towns is bad, same quadrant, but the next night we look at it and sure enough, it's a codo too. 00:14:52.620 --> 00:15:01.330 So here, just in a matter of minutes, we've been able to find 2 voucher calls for one species for one quadrant for two different nights, and it just goes very quickly. 00:15:01.640 --> 00:15:03.080 So you can do the same for big Browns. 00:15:03.250 --> 00:15:05.150 You might not find them to be depressed. 00:15:05.160 --> 00:15:12.710 I find them to be other species, then one of them turns out to be a big brown and you can go through the rest of your data like this. 00:15:13.120 --> 00:15:18.310 And this is your first pass where you can now do is mine your data for things you might have missed. 00:15:18.520 --> 00:15:23.870 So we didn't find a palette bad earlier, but we see here in the second column that there is at least one palette bad here. 00:15:23.970 --> 00:15:37.440 So if we sort on the second column alphabetically, it shows us that well, there's five files that might contain a palette that we can now look at those and particularly potentially pull a pallid bat out of one of those files to add to our our species list. 00:15:38.210 --> 00:15:43.900 Also remember we found how's bigger bets on two nights at one detector. 00:15:43.990 --> 00:15:48.000 But here in the third column, here's another potential one at a different detector. 00:15:48.010 --> 00:15:52.740 We can sort and look at that, and here's another one at the same detector. 00:15:52.750 --> 00:15:54.200 We found others, but on a different night. 00:15:54.250 --> 00:16:00.200 So basically, it's very powerful to be able to sort and mine your data. 00:16:00.410 --> 00:16:02.960 And apologies for these speed, but I know there's a lot of information. 00:16:02.970 --> 00:16:04.140 That's what recordings are for. 00:16:04.230 --> 00:16:05.010 Thank you for your time. 00:16:07.710 --> 00:16:08.190 Awesome. 00:16:08.080 --> 00:16:08.360 What? 00:16:08.200 --> 00:16:10.060 Thank you so much, Donald. 00:16:10.570 --> 00:16:13.780 So now we'll move on to Shannon Hilti. 00:16:18.260 --> 00:16:18.730 Awesome. 00:16:17.450 --> 00:16:20.550 It will share on access database and workflow. 00:16:21.530 --> 00:16:22.530 Alright, you see in that. 00:16:23.190 --> 00:16:23.310 Yes. 00:16:22.540 --> 00:16:24.650 OK, awesome. 00:16:24.660 --> 00:16:25.850 Alright, thank you. 00:16:25.860 --> 00:16:27.440 Awesome job Donald. 00:16:27.800 --> 00:16:28.490 I have everybody. 00:16:28.500 --> 00:16:34.730 My name is Shannon Hilti am the non game wildlife biologist and bat specialist with Montana Fish, Wildlife and parks. 00:16:35.620 --> 00:16:42.570 On Tuesday, you guys heard from the Montana Natural Heritage Program folks about Montana's workflow on data collection and processing. 00:16:42.660 --> 00:16:48.040 And I'm going to take it a step further and talk about manual vetting using a Microsoft Access database. 00:16:48.720 --> 00:16:55.660 So as the heritage program folks pointed out, we manage and process acoustic data for various projects including NABat. 00:16:55.670 --> 00:17:05.090 And so we house terabytes and terabytes of data and to work with this volume, it really helps to have a database that we can filter through. 00:17:05.750 --> 00:17:12.560 So Alexis introduced you to our database on Tuesday, kind of the first step for us, for Navas, figuring out what needs to be vetted. 00:17:12.680 --> 00:17:21.680 And so if we think about like a very basic goal, say we want to manually confirm as many species as we can at each detector within an A back grid cell. 00:17:22.520 --> 00:17:25.590 So we have various tables within our database. 00:17:25.860 --> 00:17:27.280 You can see these over here on the left. 00:17:28.400 --> 00:17:39.970 These this is just a small snippet of the actual database, but each NA backgrid cell corresponds to a site ID in our database and then each detector within the cell corresponds to one of these locations. 00:17:40.060 --> 00:17:48.010 This table also includes sahabat output, kaleidoscope output, as well as information associated with each bat sequence. 00:17:48.100 --> 00:17:54.650 So this is the small snippet there's I think 155 fields in this particular table, but that gives you kind of an idea of what we're looking at. 00:17:55.190 --> 00:18:05.180 So one of the great things about using an access database is that we can write queries or use filters to pull out information that we want to look at from this massive pile of data. 00:18:05.570 --> 00:18:21.240 So I just wrote a very simple query to kind of give me a summary on data collected at grid cell 3784, detector four during the 2021 field season and you can see in this summary that's on about auto ID 351 call sequence to a species. 00:18:21.470 --> 00:18:26.800 The rest were labeled undetermined and the hand class field in our database. 00:18:27.190 --> 00:18:29.410 Umm, there are lots of values that correspond to this. 00:18:29.420 --> 00:18:40.160 I don't have time to talk about all of them, but we use AD to confirm that we have definitively identified as sequence P is probable and then X means that we've actually completed the survey for that particular species. 00:18:40.590 --> 00:18:46.840 I'll talk about that in a second, but here you can see that we had one definitive palette bat that we already identified. 00:18:46.970 --> 00:18:50.640 And so the rest of the calls that have been added by saying about, we don't even need to look at. 00:18:50.650 --> 00:18:57.960 So we ex them out so that they don't show up in queries that are running in the background and you can pull this kind of information in access using filters. 00:18:58.240 --> 00:19:07.710 I just prefer queries myself, so the next step is to actually go through and ID the other species that we haven't classified a definitive call for. 00:19:08.060 --> 00:19:11.970 So Alexis is the person that has to manually vet everything right now for the most part. 00:19:11.980 --> 00:19:14.390 And again, we're dealing with millions of bat sequences. 00:19:14.400 --> 00:19:18.530 And so she can't look at every single call that we have from those surveys. 00:19:18.730 --> 00:19:28.220 And so the next step is to filter out the sequences with the highest likelihood of being the species of interest based on parameters in our Montana back key. 00:19:28.790 --> 00:19:33.980 And we developed this key for Montana species based on several other keys. 00:19:34.150 --> 00:19:35.000 Joe Sussex. 00:19:35.010 --> 00:19:39.060 Professional opinions and then we've modified it based on what we've observed in the state. 00:19:39.540 --> 00:19:52.700 And so this along with standardized site selection and detector deployment methods really helps us standardize the way that we are making decisions while we're manually vetting, which as Donald pointed out, is important. 00:19:53.250 --> 00:19:59.030 And so this helps us filter out sequences that have the highest likelihood of being the species of interest. 00:19:59.520 --> 00:20:05.630 And there are actually 3 pages to this document, one the one page you can't see goes over the workflow that Donald kind of outlined. 00:20:05.820 --> 00:20:09.930 But as you can see, we've got our important parameters around the top here. 00:20:10.360 --> 00:20:14.470 So if we go back to our example, let's say we wanna tackle hoary bat. 00:20:14.480 --> 00:20:21.030 Now the query that I wrote pulled out 43 sequences that were autoid during as Lacey during this survey. 00:20:21.100 --> 00:20:23.850 And so here's the section of the key that pertains to Lacey. 00:20:24.190 --> 00:20:26.700 Umm, this hand class priorities column? 00:20:26.710 --> 00:20:36.560 Essentially, it's our first pass, so the calls have to meet these minimum criteria to move forward in terms of having a human look at them, they don't meet these criteria. 00:20:36.570 --> 00:20:40.120 We give them a Z in our database and then we just don't even look at them. 00:20:40.630 --> 00:20:44.910 So you can see 19 we're classified as Z in this particular example. 00:20:45.210 --> 00:20:52.760 The bold writing under this diagnostic characteristics that gives us the actual things that we need to see in a call sequence to call it definitive. 00:20:52.770 --> 00:20:54.570 And so there are 34 sequences here. 00:20:54.910 --> 00:20:57.310 I'm sorry, 24 that meet those requirements. 00:20:57.320 --> 00:21:02.070 So again, we filter out everything that had FC greater than 20. 00:21:02.180 --> 00:21:14.870 So we're left with the lower calls and then for horseback in particular, I like to sort on duration and I look at the longest calls first because in my experience that has a lot of success in finding definitive calls. 00:21:15.910 --> 00:21:22.970 But once you find one, you exit out, you're done with that species, and you can move on to the next and work through the key using those various filters. 00:21:23.360 --> 00:21:25.290 And I know that the town to throw at you. 00:21:25.300 --> 00:21:27.480 It's kind of like you dating for back calls. 00:21:28.170 --> 00:21:31.880 You know, I would say that these tricks worked really well for us in dealing with large data sets. 00:21:31.890 --> 00:21:34.460 I think they work on smaller datasets as well. 00:21:34.910 --> 00:21:37.360 If you have questions, feel free to reach out to any of us. 00:21:37.370 --> 00:21:40.180 I also want to give Brayden a shout out because he's the brain behind this. 00:21:40.190 --> 00:21:41.460 He created the database. 00:21:41.470 --> 00:21:44.480 He does all of the wizardry, so let us know if you have questions. 00:21:46.850 --> 00:21:47.620 Wonderful. 00:21:47.630 --> 00:21:50.630 Thanks for taking us through that, Shannon. 00:21:50.640 --> 00:21:55.210 And so now we'll move over to Ted, who will talk about Echo clean. 00:21:56.960 --> 00:21:59.230 We'll just confirming you can see my screen and hear me. 00:22:00.120 --> 00:22:00.260 Yes. 00:22:01.260 --> 00:22:01.590 Yeah. 00:22:01.600 --> 00:22:01.930 Good. 00:22:02.040 --> 00:22:05.050 OK, so those are perfect introductions to what I'm talking about. 00:22:05.060 --> 00:22:09.010 It sounds like we all have really similar kind of workflows, so this is another tool. 00:22:09.160 --> 00:22:10.120 It's called echo clean. 00:22:10.130 --> 00:22:12.950 I'm just gonna walk through it rather than have a presentation. 00:22:12.960 --> 00:22:17.610 So as we kind of heard, we're taught about output. 00:22:17.620 --> 00:22:20.050 You might get a situation like this where thought about things. 00:22:20.060 --> 00:22:20.410 It's. 00:22:20.540 --> 00:22:22.910 It's mica, my oldest californicus. 00:22:22.920 --> 00:22:24.210 It's 15 out of 16. 00:22:24.220 --> 00:22:26.650 It's not confused with any other species you might like. 00:22:26.660 --> 00:22:27.070 Think. 00:22:27.170 --> 00:22:29.150 Hey, we wanna trust song about on that. 00:22:29.240 --> 00:22:33.760 We have other cases where, umm, there's not very many good pulses. 00:22:33.770 --> 00:22:35.720 It's confused between two species. 00:22:35.870 --> 00:22:39.020 We may say, why would I ever want to ever wanna look at this? 00:22:39.030 --> 00:22:42.980 So the tool that we have called ECHO Clean is a is a Python script. 00:22:42.990 --> 00:22:46.780 It's meant for situations where different from mNABat. 00:22:46.790 --> 00:22:51.680 You wanna assign an ID to every single file in your in your data set. 00:22:51.690 --> 00:22:54.960 So identify species and assign things to categories. 00:22:55.030 --> 00:22:59.220 So just to be really risky, I'm gonna try to run this thing. 00:22:59.910 --> 00:23:02.410 Umm, ohh, uh. 00:23:02.800 --> 00:23:06.420 I'm gonna try to run this thing live here and show you that. 00:23:06.430 --> 00:23:07.370 How easy it is? 00:23:07.380 --> 00:23:15.280 Like I said, Python script and maybe you're scared by that, but I'm gonna say that once you get it set up, it's easy to run it. 00:23:15.290 --> 00:23:18.380 It's like you tell it the directory that you're going to. 00:23:18.690 --> 00:23:21.360 You type in a single line of code that says. 00:23:21.370 --> 00:23:26.660 What rules do I wanna use and what data set do I wanna use that comes from sauna? 00:23:26.670 --> 00:23:30.120 Bat the rules have to be in the same directory as or. 00:23:30.130 --> 00:23:32.570 I'm sorry the data set has to be in the same directory as the rules. 00:23:33.100 --> 00:23:37.150 When we press this button it then says icon multiple sheets. 00:23:37.160 --> 00:23:38.340 I'll get into this later. 00:23:38.460 --> 00:23:39.910 Which rules do I wanna use? 00:23:39.920 --> 00:23:45.670 I'm gonna say I'm gonna use nine and it's gonna classify so this data set I had Central Coast. 00:23:45.720 --> 00:23:51.440 It was around like 2200 passes that had the song about output from there. 00:23:51.560 --> 00:23:53.730 So right now it's kind of chugging through. 00:23:53.820 --> 00:23:56.550 This one takes a little bit longer than I'm. 00:23:56.660 --> 00:23:59.540 I'm hoping that it'll it'll still be good here in a second. 00:24:00.680 --> 00:24:01.350 Here we go. 00:24:01.400 --> 00:24:08.090 So look, it evaluated 22120, our 2222 passes 220 of those matched. 00:24:08.100 --> 00:24:15.410 One of our Rule 2 did not match any rules and then so that gets us an output that looks like this. 00:24:15.420 --> 00:24:37.160 Like we've seen something like this before, so at the top again we're sorted by the 1st like Donald was explaining which which of these are possible ampo which is these are possible effort which of these are possible label then it says should I inspect this one it says no you don't need to inspect that because it was ample 11 of 12 sound about was really really unconfident about that. 00:24:37.650 --> 00:24:38.680 How about the next one? 00:24:38.690 --> 00:24:44.180 Well, it's not about it first thought it was AMPA, but ended up assigning it to Tabor and it was seven out of 16. 00:24:44.190 --> 00:24:48.800 So if you want to assign an ID to this, whether it be a category or a species, yes you wanna. 00:24:49.120 --> 00:24:52.220 Yes, you wanna go ahead and look at that. 00:24:52.590 --> 00:24:53.020 OK. 00:24:53.030 --> 00:24:57.780 So then what do I mean by the the rules? 00:24:57.790 --> 00:25:04.450 So these are a set of rules that I come up with for various sites based on my experience with a lot of manual vetting first. 00:25:04.460 --> 00:25:13.290 So I said I looked at this and I said wow, so many times when I manually vetted this, if it was greater than nine out of nine for AMPA, it was always AMPA. 00:25:13.300 --> 00:25:22.290 So we don't need to look at that one anymore versus when it was six out of nine and it was confused with Lano or Lacey or Tabor, et cetera. 00:25:22.690 --> 00:25:26.950 That hardly ever resulted in an idea of any of those species, so let's not even look at that. 00:25:26.960 --> 00:25:27.980 Let's put that in the queue. 00:25:27.990 --> 00:25:31.930 25K category or the 25K category. 00:25:33.710 --> 00:25:40.300 So these are rules that just come up at came up out of my out of my head from previous experience vetting these calls. 00:25:40.510 --> 00:25:41.580 We're working on a process. 00:25:41.590 --> 00:25:51.300 We're actually have some quantitative things, so instead of it being 6 out of nine, it would also include things like the characteristic frequency was less than 27.32 or something like that, right? 00:25:51.310 --> 00:25:52.640 We can get to that later. 00:25:52.710 --> 00:25:58.740 Also, these rules are made for individual sites, so I have different, uhm. 00:25:58.900 --> 00:26:02.320 A different rule sets for different places that we work. 00:26:02.550 --> 00:26:03.100 What else? 00:26:03.550 --> 00:26:07.330 So we can see for AMPA, there's a lot of rules for codo. 00:26:07.630 --> 00:26:12.600 Basically, it's a rare species and when we see it, we wanna just when it comes up, we wanna look at most of those. 00:26:12.610 --> 00:26:13.740 So there's not very many rules. 00:26:13.750 --> 00:26:20.760 Basically, it's like if there's greater than three out of three, we want to look at that Ebola ton of rules spotted that massive bats. 00:26:20.770 --> 00:26:31.010 Basically, if sonnet ever suggests that we want to look at that, that's gets to Donald's or concerned about what do we do about rare species that's I'm covers that again going fast here. 00:26:31.700 --> 00:26:34.030 Umm, so yeah, this is just a tool. 00:26:34.040 --> 00:26:40.350 So again, when you get these results, so let's say it says no inspect, but then you didn't find any other amphos in your data set. 00:26:40.560 --> 00:26:41.310 It's a tool. 00:26:41.320 --> 00:26:47.150 You can override it and still go in and assign a manual ID and just say you change your inspect to yes, we've also. 00:26:48.940 --> 00:26:54.680 It works for Kaleidoscope data also, so there's kaleidoscope rules using different parameters and kaleidoscope results. 00:26:54.690 --> 00:26:58.670 So you can use those too, and I'm just gonna go quickly to you. 00:27:00.440 --> 00:27:04.690 Uh, this one and I'll put this link in the chat here. 00:27:04.700 --> 00:27:09.530 So this is on GitHub, my partner in in crime on for a lot of things. 00:27:09.540 --> 00:27:12.670 Brendan Ward haven't helped me develop this and put it on his GitHub here. 00:27:13.600 --> 00:27:16.550 It explains how to install Python, which is actually the hardest part. 00:27:16.560 --> 00:27:21.510 The running of it you saw was insanely easy, and then there's down here in the bottom. 00:27:21.520 --> 00:27:22.790 We just added yesterday. 00:27:23.100 --> 00:27:26.370 There's some example rules that get you started like here's my excel sheet. 00:27:26.380 --> 00:27:28.090 These are gonna get you started, et cetera. 00:27:28.100 --> 00:27:31.140 So I'll end up there and put this in the chat and we can move on. 00:27:32.070 --> 00:27:33.560 Thanks so much, Ted. 00:27:33.260 --> 00:27:34.360 Yeah, yeah. 00:27:33.570 --> 00:27:36.320 So now we're gonna move on to to Ben. 00:27:36.330 --> 00:27:40.220 Who's gonna share Kaleidoscope and Annabelle or Anna look? 00:27:41.980 --> 00:27:43.480 All right, let me get this up. 00:27:47.090 --> 00:27:48.230 OK, see my slides. 00:27:51.200 --> 00:27:51.400 Yep. 00:27:51.830 --> 00:27:52.560 All right, cool. 00:27:52.670 --> 00:27:56.720 So yeah, Donald had a really good intro on the terms. 00:27:56.890 --> 00:28:05.340 When you're manually setting, especially using sauna, bat and not all of them are available in Kaleidoscope, but quite a few of them are. 00:28:05.350 --> 00:28:08.290 Some of them have different terms, so I'm just gonna go through those. 00:28:10.180 --> 00:28:20.760 First off, when you after you do an audio ID in kaleidoscope, you get some output files in the folder that you've chosen for the output directory and those are here under the underneath the folders. 00:28:20.770 --> 00:28:30.970 In this little screen shot, some of them aren't really too useful for manual wedding, but the ID the two ID files are very useful, so we'll take a look at those first. 00:28:30.980 --> 00:28:33.470 Is the ID summary dot CSV file. 00:28:33.480 --> 00:28:38.960 You can open this in Excel and you can see a couple tables. 00:28:38.970 --> 00:28:56.150 The first one that you're looking at is the calls per species by night, so it shows the number of calls that were auto ID to each species that you chose in your list of species and across the top is the total number of calls across all nights in your data set for each species. 00:28:56.990 --> 00:29:01.840 And next to that is another table showing the presence P values for each species. 00:29:01.850 --> 00:29:08.360 So the lower lower the value, the higher probability that the classification was correct. 00:29:08.990 --> 00:29:13.520 Of course, it does need to be manually vetted, but you can take a look at these here and across the top row. 00:29:13.530 --> 00:29:21.460 Again, is are the probabilities across all the nights, so the next file is the ID CSB. 00:29:21.670 --> 00:29:40.970 This is a large, pretty large, usually depending on how many files you have, it has a lot of columns as well as every file in your data set, and it has the column measures or metrics that are used in the analysis and from my experience, this is sorted by auto ID and then by quality kind of based on some of the columns. 00:29:40.980 --> 00:29:53.460 I don't know exactly which columns they're sorting by, but usually the higher quality calls are at the top, so so just some simple terms that are in this first I'll go through those out files easy. 00:29:53.460 --> 00:29:55.260 Is just the name of the output file. 00:29:55.820 --> 00:30:12.240 Date minus 12 is the timestamp of the file minus 12 hours, so that's how you can get your night, your survey night from Collider scope, output audio ID is the automated classification ID that was given and the manual ID is where you'll be putting your manual ID or no ID. 00:30:14.140 --> 00:30:27.990 So to zoom in here on a couple of the columns that are kind of similar to what you've seen in song about on the left side here we have the auto ID that was given to the species and then pulses are the total number of pulses that were identified to. 00:30:28.140 --> 00:30:29.550 Sorry, I'm skipping ahead a little bit. 00:30:29.640 --> 00:30:36.710 Alternate one and alternate two are kind of similar to the 1st and 2nd, 3rd and 4th that we're in. 00:30:36.720 --> 00:30:37.530 Sonipat. 00:30:37.640 --> 00:30:43.640 So these are other species that are potential for potential ID for the file. 00:30:44.000 --> 00:30:50.650 Sometimes it could be multiple species recorded in the same file, or it could be just an alternate possibility. 00:30:50.880 --> 00:30:58.600 So you can take a look at those and then we have pulses, which is the total number of pulses that were identified to any species. 00:30:58.610 --> 00:31:00.420 So it could be those other alternates in there. 00:31:01.030 --> 00:31:01.470 Umm. 00:31:02.090 --> 00:31:05.880 And then next to that is matching and match race ratio. 00:31:05.890 --> 00:31:14.800 So matching are the number of pulses that were ID to the audio ID species and the match ratio is the number of matching over the number of pulses. 00:31:14.810 --> 00:31:22.640 So you can kind of look at this match ratio as a percentage of the pulses that are ID to that species and it's sort of a quality metric. 00:31:24.470 --> 00:31:37.420 Also, margin is something that you can read the Kaleidoscope manual to get more information about all of these, but they list this as sort of within species score for the call classification. 00:31:38.270 --> 00:31:40.920 I'm kind of trying to look into this as my script. 00:31:40.930 --> 00:31:51.140 It all show you in a little bit, but I'm not too sure about this as an actual measure of quality, so some more metrics here. 00:31:51.380 --> 00:31:53.610 FC is characteristic frequency. 00:31:54.480 --> 00:31:56.100 I think Donald mentioned that earlier. 00:31:56.700 --> 00:32:05.820 Uh, but this is one that you can look at, and if you have those tables that give you like a range of values that are common for each species, this is something you can look at to compare. 00:32:06.500 --> 00:32:12.190 Umm, and then Fmax and Fmin are the high frequency and low frequency. 00:32:12.200 --> 00:32:15.370 And this is across all pulses in the in the recordings. 00:32:15.380 --> 00:32:26.900 So it's an average and these are also valuable for if you have those tables and you know what to look for for each species, it's a little bit easier to think about these if you're looking at a call. 00:32:26.910 --> 00:32:36.350 So it opened one up in analog here and it you can see on the right side it's showing the values for these different things characteristic frequency Max and min frequency. 00:32:37.340 --> 00:32:39.130 Uh, and you can see them there. 00:32:39.140 --> 00:32:42.570 The numbers are lined up and kilohertz on the left side. 00:32:43.260 --> 00:32:45.470 You can explore that a little more when you have time. 00:32:47.530 --> 00:32:50.040 Umm, another one that's in here is characteristic slope. 00:32:50.050 --> 00:33:00.230 I don't know if a lot of people use this like actual value, but it's something you can look at when you're looking at the files visually represented here to get an idea of the call shape. 00:33:00.240 --> 00:33:15.460 That's something you kind of key in on when you're vetting and duration is the average duration of pulses in the file, which I believe is also in most of those tables that give you values to look for when you're manually betting. 00:33:17.470 --> 00:33:20.380 So the 2018 guide to processing. 00:33:21.170 --> 00:33:21.610 Umm. 00:33:22.230 --> 00:33:23.820 Document that was published. 00:33:24.190 --> 00:33:26.540 I believe it's part of the NABat central office. 00:33:27.110 --> 00:33:28.440 You can find that online. 00:33:28.950 --> 00:33:38.120 It recommends sorting for Kaleidoscope sorting by folder, matching date and auto ID as kind of a workflow for vetting. 00:33:38.130 --> 00:33:40.890 So you're not looking at a bunch of files that are lower quality? 00:33:41.560 --> 00:33:42.060 Umm. 00:33:42.420 --> 00:33:47.920 And then it recommends reviewing that match ratio and then you put the species that ID in the manual ID field. 00:33:48.650 --> 00:34:00.490 Umm, so I'm going to move on to my next part, which is a call quality sorting for analog W or any other purpose I I've called it this but you could use it for other software. 00:34:00.500 --> 00:34:08.630 If you look at the zero crossing files and something else, and I wrote this in Python And it uses the metrics from Kaleidoscope Auto ID. 00:34:08.640 --> 00:34:10.290 So I'm gonna open this here. 00:34:10.920 --> 00:34:25.480 So first we have here the analog window and umm, you know was it doesn't, it's pretty old software and it's kind of lacking some features that are very useful for manual vetting you can. 00:34:26.660 --> 00:34:32.450 Disperse the files into folders by species, but it doesn't sort them by quality. 00:34:32.740 --> 00:34:36.450 If you have insight which is our new software, you can sort by quality. 00:34:36.840 --> 00:34:47.130 I'm in other things that you want to sort by, but a lot of people learn to use analog W and are still using it, so they're just familiar with it in like the workflow in it. 00:34:47.420 --> 00:34:55.010 And I've just written some code to kind of enhance that to tailor it to vetting for NABat data. 00:34:55.060 --> 00:34:58.540 So take a look here at. 00:35:00.210 --> 00:35:01.110 I don't see where am I. 00:35:06.260 --> 00:35:24.950 Python script that I wrote and this is just it's a pretty long script so I'm not going to show you a lot of it that I wanted to Scroll down to one section here so it gives people options as they're running it and if umm, the person chose that they wanted to sort by those call metrics, it's going to read the file. 00:35:25.840 --> 00:35:26.300 Umm. 00:35:27.220 --> 00:35:29.070 And I've opened one in a hex editor. 00:35:29.080 --> 00:35:36.720 Here you can see at the very beginning there's some header info, but I found it was easier to get this in the script by looking for the. 00:35:38.500 --> 00:35:42.570 ID value here and give grabbing that you can see the score is in here. 00:35:42.580 --> 00:35:43.380 That's the margin. 00:35:43.390 --> 00:35:44.670 The pulses are in here. 00:35:45.280 --> 00:35:48.310 Number matching are in here and alternates are here. 00:35:48.320 --> 00:35:53.520 So this one had, it was a lazy, but it was given also free tailed as an alternate. 00:35:53.920 --> 00:35:57.450 So this code reads all of that information and someone chooses to sort it. 00:35:57.460 --> 00:35:57.930 It will. 00:35:58.410 --> 00:36:04.360 And this is kind of if you can make sense of this is showing how each thing is sorted. 00:36:04.370 --> 00:36:16.940 So I'm first sorting by path and it's ascending species ascending number of pulses matching are descending pulses descending margin descending and so this is the order of sorting here. 00:36:17.330 --> 00:36:20.640 So I'm gonna open this and sort some data. 00:36:30.430 --> 00:36:30.660 Maybe. 00:36:37.460 --> 00:36:44.370 So as I mentioned, yes or two days ago I tried to write my code so that people can just run it on there and without having to know Python. 00:36:44.380 --> 00:36:58.720 So if I give it to you and you install Python And it's working correctly, you don't need to look at any of this and you can just double click the Python file and it will pop up a window in command prompt, give you some information and introduction. 00:36:58.730 --> 00:36:59.640 I'll zoom in a little bit here. 00:37:00.390 --> 00:37:19.060 Umm, so it renames files and I just want people to be aware of that because that's how it has to work in order for the sorting to happen in analog it can be undone, but I just want people to know that the files are gonna be renamed, so you have to agree to that first and then I'm going to do this path here, which would take a look. 00:37:19.070 --> 00:37:28.080 It has files and the function of nights and some noise files and things, so I will paste the path in here and it's asking do I want to search recursively? 00:37:28.090 --> 00:37:33.730 Yes, because it's night folders, nightly folders and do I want to sort files or unsought them. 00:37:33.740 --> 00:37:40.020 So I'm sorting umm and it gives you the option of grouping by site or by night. 00:37:40.150 --> 00:37:47.120 I'm just gonna do it by site here so that we can just look at all the files sorted, have more files that way to look at. 00:37:47.770 --> 00:37:55.730 Do you want to group calls given no ID and have an alternate species so you can ignore them if you want to, or you can include them. 00:37:56.070 --> 00:37:59.810 I want to include them in this case so we can see how the alternate ID sorting works. 00:38:00.570 --> 00:38:01.160 Umm. 00:38:01.970 --> 00:38:07.380 And I do want it to use the metrics to sort them, so I'm agreeing to that and now it's running. 00:38:07.670 --> 00:38:15.000 It's really quick and it found 163 files with species Auto IDs and sorted those and that's it. 00:38:15.310 --> 00:38:28.860 So we can look in this folder here now and since I told it to do it by site and move them out of the nightly folders and put them all here, it's renamed them and given them their species IDs and has alternates in here and it has this little. 00:38:28.870 --> 00:38:36.210 This 00010 gives you 3, so that's so that it sorts it in order by call quality based on some of those metrics I had. 00:38:36.420 --> 00:38:39.750 So we can drag them into here and look at them. 00:38:39.760 --> 00:38:46.820 You can see this is a potential EPP group call with a lot of pulses, and I'm just gonna go through them and it's you can see the file name up here. 00:38:46.830 --> 00:38:57.450 If that's not too small and I'm just gonna go through down in quality basically and you can see that they're a little noisier or fewer pulses or maybe a little weird looking. 00:38:57.460 --> 00:39:03.660 So we're going through the app through quality, and now we're into Lacey. 00:39:03.930 --> 00:39:06.760 So there's more pulses and it looks like a decent call. 00:39:07.450 --> 00:39:08.960 This is a kind of weird call. 00:39:08.970 --> 00:39:15.930 There aren't a lot of lazy calls in here I think, but this and I'm open to other suggestions. 00:39:16.010 --> 00:39:16.500 Am I out of time? 00:39:21.650 --> 00:39:22.010 That's OK. 00:39:14.960 --> 00:39:23.490 So, so, so Ben, I'm gonna uh, this is really wonderful, but I'm gonna cut you off, and we're gonna move on to Jason. 00:39:23.730 --> 00:39:24.080 All right. 00:39:23.760 --> 00:39:26.980 Who's gonna share in about insight workflow with us? 00:39:29.940 --> 00:39:30.210 OK. 00:39:30.220 --> 00:39:30.340 Yeah. 00:39:29.940 --> 00:39:30.760 Thank you so much. 00:39:31.750 --> 00:39:32.440 Uh, so thank you. 00:39:32.450 --> 00:39:33.370 My name is Jason Ray. 00:39:33.380 --> 00:39:39.170 I'm working for wildlife conservationist, Society Canada, and we've been coordinating coordinating BC and about sampling since 2016. 00:39:40.280 --> 00:39:40.700 Umm. 00:39:41.500 --> 00:39:44.070 He'll be talking about a basic and about insight workflow. 00:39:44.080 --> 00:39:45.530 While I share my screen. 00:39:50.120 --> 00:39:56.440 And I'll be using some of the same terms that everybody else has used, so we'll skip the definitions of those just to kind of try and save some time here. 00:39:59.960 --> 00:40:00.180 OK. 00:40:02.630 --> 00:40:04.760 So anybody inside is developed by Titley Scientific. 00:40:04.930 --> 00:40:13.460 It's a free to use to analyze files recorded on their NABat detectors, which are swift Walkabout scout for the Express with updated firmware. 00:40:14.470 --> 00:40:27.270 If you want to use other brands of files from other brands of detectors, you the charge a one time unlock fee of 400 USD and we typically use this to manually analyze a bunch of the files in PC. Here. 00:40:27.360 --> 00:40:28.830 So quick summary of initial steps. 00:40:28.880 --> 00:40:39.650 To organize our files into folders, we like to do preprocessing with Collada Scope pro and SauNABath both to assign initial auto IDs and sort files into possible species for manual analysis as well as noise screening. 00:40:40.720 --> 00:40:49.880 We do our manual analysis and we follow up with the second step of vetting suspicious I's unexpected species and trying to pull up missing species by second analyst. 00:40:51.100 --> 00:40:52.850 Then we summarize for statistics on importing. 00:40:55.320 --> 00:40:56.900 This is the first screen of about insight. 00:40:57.220 --> 00:41:02.390 I'll just give a quick orientation starting from the top left, the color histogram up there can be used to adjust brightness. 00:41:02.400 --> 00:41:04.450 Contrast spectral intensity of the spectrogram. 00:41:05.100 --> 00:41:08.820 Adjusting these can be helpful to reduce visibility of noise in the file and help highlight that pulses. 00:41:10.420 --> 00:41:12.750 The stores window summarizes the files you've loaded into the program. 00:41:14.470 --> 00:41:34.910 The bottom left one is a lot more functionality than I have time to talk about, but it either displays marks, marked recordings let you record and edit filters that will screen up pulses and each recording by specific parameters, or search and display files that fit certain criteria or filters and BC we're actually analyzing all files that receive an auto ID label because we're looking at using the activity data as well. 00:41:35.280 --> 00:41:40.050 But these filters are one of the better tools you have to start sorting your recordings into higher quality files. 00:41:40.060 --> 00:41:42.420 That will be your most likely to pull out a specific idea. 00:41:42.430 --> 00:41:43.270 You're looking for as well. 00:41:45.370 --> 00:41:54.100 At the bottom of the window you see my little sets here and bottom right is the tab that will show the metadata within the file that you're viewing now. 00:41:54.110 --> 00:42:09.880 Because we've put both Collado scope and sauNABat IDs in here that they're gonna show up in this metadata field here for that other opinion that we can use while we manually that last, we have the top right window that displays the zero cross parameters of all pulses currently visible in the spectrogram. 00:42:09.890 --> 00:42:11.740 And you'll see that more than in a second. 00:42:11.950 --> 00:42:14.770 And then of course the spectrogram here shows up in the big main window. 00:42:15.820 --> 00:42:21.880 So now that we're oriented, you hit file and open directory to load a folder of recordings into your store, I'd select. 00:42:21.890 --> 00:42:30.650 I would suggest loading a grid cell or similar grid cells with similar species sets at one time, so you can kind of keep that search image in your head. 00:42:32.010 --> 00:42:40.460 Uh hotkeys from moving between files are the same as analyst analytics or analytic W so if you're familiar with those, it's pretty quick to jump in here. 00:42:41.330 --> 00:42:52.570 Square Back is to move back and forth between files in your store and ASD and F keys to move forward in time in the currently loaded file and you just double click on the item in the store to open that recording. 00:42:54.480 --> 00:43:02.330 So you can also see the time scales of the top F1 to F-10, the compressed and true time button which has a hockey of spacebar to toggle between them. 00:43:03.820 --> 00:43:07.310 Our full spectrum files you can talk between full spectrum and 0 cross representations. 00:43:07.580 --> 00:43:10.950 I like to use the overlay because it gives me the most information in the most compact form. 00:43:12.700 --> 00:43:14.530 You can adjust the zero crossing threshold. 00:43:14.540 --> 00:43:16.930 That insight uses to detect pulses with this slider. 00:43:17.100 --> 00:43:19.650 Typically I have found the auto FS works fairly reasonably. 00:43:20.890 --> 00:43:22.220 You can adjust the FFT. 00:43:22.290 --> 00:43:22.580 Uh. 00:43:22.590 --> 00:43:28.480 The playback the audio using various different methods and lastly the media on trigger frequencies for zero crossing thresholds. 00:43:28.490 --> 00:43:35.040 We typically leave the leave these at default, but if you have a specific value you're looking at, you can adjust that to screen out a little bit more of the noise. 00:43:35.050 --> 00:43:36.070 That's not the bat you're looking. 00:43:37.580 --> 00:43:43.860 Umm, next step is to adjust the store window to group the files by folder and auto ID. 00:43:44.990 --> 00:43:53.890 This is where you can also sort your files by metrics that other IDs have put in there or sort your files by name or date or whatever else you wish. 00:43:56.800 --> 00:43:57.190 Are you? 00:43:57.240 --> 00:44:07.130 Like I said, I typically suggest grouping by auto ID and location for for our workflow, but if you're using some other priority metrics like you want to look at, the most likely species files first. 00:44:07.140 --> 00:44:07.960 You can do that here as well. 00:44:13.170 --> 00:44:15.020 Next, you set up some species lists. 00:44:15.580 --> 00:44:18.310 That's just using the species list option on the top left. 00:44:20.040 --> 00:44:23.190 Each label gets their own row and you can have as many lists as you would like. 00:44:25.590 --> 00:44:29.860 Means 0 cross metrics of all pulses are on screen are shown in the top right. 00:44:29.870 --> 00:44:36.250 As I said, these are the things like the SC, the duration, the FFC and stuff that Ben had already previously talked about. 00:44:37.100 --> 00:44:44.680 What I like about insight is as well it gives you the option to show the metrics of an individual pulse on your screen just by most mousing over it. 00:44:44.940 --> 00:44:46.700 So that's this window here in the middle. 00:44:47.590 --> 00:44:54.570 You can also create your own metrics as well that are combinations of the above SC and FC kind of thing. 00:44:55.840 --> 00:45:00.640 So once you've looked at your file, you've found pulses that you think are indicative of of a specific species. 00:45:00.650 --> 00:45:09.720 You would move down to your metadata lists here, hit the radio button to select one that fits for the frequency category they're using. 00:45:10.500 --> 00:45:10.910 Umm. 00:45:12.190 --> 00:45:22.920 I've organized my species sets into broad frequency categories like you see here, high, medium, and low, so I don't have to switch between the very often and then you can use the numpad on the right of your keyboard to assign labels one to 10. 00:45:23.770 --> 00:45:30.200 I've typically kept my species sets at less than 10, so I can only use the number pads and just switch between species sets find. 00:45:30.210 --> 00:45:32.590 I find that the smoothest workflow for me. 00:45:34.670 --> 00:45:40.820 You can then manually edit the information in certain fields as well, and about insight does lock. Most of these fields. 00:45:40.830 --> 00:45:45.310 But things like manual ID are also manually editable, so you don't need to use these buttons. 00:45:45.320 --> 00:45:48.130 You can just kind of double click on that field and enter your text there too. 00:45:50.570 --> 00:45:58.020 And while applying labels we do like to look at the full sequences of pulses, including parameters and call shapes using both Glatopa Pro and some about labels too. 00:45:58.030 --> 00:46:01.450 So those are also available in this metadata viewer in the bottom right. 00:46:05.410 --> 00:46:11.150 And once you've applied your label, move on to the next file with the square bracket, so those are the basics that just plan to talk about. 00:46:11.620 --> 00:46:12.850 Thanks for paying attention. 00:46:14.110 --> 00:46:16.120 There are many other options that I didn't cover here. 00:46:16.230 --> 00:46:25.440 Inside does not provide its own auto ID, but it does provide a nice, capable and concise information summary of the other autoid programs. 00:46:26.420 --> 00:46:30.930 And we use this to, uh, many leave at most of our calls in BC here. 00:46:31.290 --> 00:46:31.780 So thank you. 00:46:33.560 --> 00:46:33.950 Great. 00:46:33.960 --> 00:46:46.870 Thanks so much, Jason, and thanks to all the panelists for for sharing and also respecting the time of others to keep us moving along, knowing that we have a really tight schedule. 00:46:47.520 --> 00:46:53.810 So we have just a few short minutes for questions and we're gonna try to stay on schedule. 00:46:53.820 --> 00:46:59.810 So we can have a greater bulk of time at the end to really dive into some of these questions. 00:46:59.820 --> 00:47:01.430 So does anybody have any questions? 00:47:01.440 --> 00:47:02.890 They wanted to pop up and share. 00:47:04.050 --> 00:47:08.560 I do notice that Donald had one for Jason. 00:47:09.430 --> 00:47:17.200 Are you looking for agreement between Sanobar and Kaleidoscope for prioritizing what species you manually vet? 00:47:20.040 --> 00:47:29.450 So that's definitely something that we would use in the manual betting step, but because we're analyzing all files, uh, we do still just go through all files more or less. 00:47:29.460 --> 00:47:36.120 We would probably look a little bit less, I guess deeply on files that do have agreement though, so, yeah. 00:47:39.130 --> 00:47:39.520 Great. 00:47:39.530 --> 00:47:40.080 Thank you. 00:47:40.090 --> 00:47:42.700 And does anyone else out there have any questions? 00:47:48.960 --> 00:47:52.550 So Ted, if you could just jump in real quickly. 00:47:52.560 --> 00:47:58.690 I know you had shared some information in the chat about the rule sets that you developed for the Echo clean. 00:48:03.000 --> 00:48:05.810 But yeah, yeah, sure. 00:48:05.820 --> 00:48:06.440 The I'm Donald. 00:48:06.450 --> 00:48:09.360 Just had a question about how portable they are from site to site. 00:48:09.370 --> 00:48:20.000 So I just wanted to make the uh, make it clear that, like the more specific you can make it to us to a general site like the more effective your rule sets are gonna be and that takes some iterative time. 00:48:20.010 --> 00:48:25.620 So like I was saying, we like when we moved, we searched first, started doing work in Wisconsin, we had a Wisconsin rule set. 00:48:25.630 --> 00:48:31.580 But then we found Ohh well North and South Wisconsin are quite different based on the relative abundance of particular species. 00:48:31.790 --> 00:48:41.730 So then we would tweak the rule sets that way, after first applying the Wisconsin data set, then seeing how many files we had to look at, and then saying so. 00:48:41.740 --> 00:48:46.050 That worked really well in northern Wisconsin, but for southern Wisconsin, we wanna tweak those things. 00:48:46.060 --> 00:49:01.230 So generally when I do it, I try to I shoot for three to 5% of the files that I'm putting eyes on and I think anyone who does this a lot of this work is realizing that there's so many files that we could never assign to species based on manual vetting, etcetera. 00:49:01.240 --> 00:49:05.390 So that's one of the big things that echo clean does, just put them in categories right away. 00:49:05.400 --> 00:49:15.800 So you don't have to spend time on those spending time on the ones that are critical for your particular area like endangered species or ones where there's confusion among a particular species. 00:49:15.810 --> 00:49:19.080 So I'm happy to answer any other questions about it. 00:49:19.090 --> 00:49:20.140 People can contact me. 00:49:20.430 --> 00:49:20.800 There are. 00:49:20.810 --> 00:49:28.740 There is an example rule set that we put up the GitHub that'll that really only works like the closer you are to Northern California than more applicable it'll be, but we have. 00:49:28.750 --> 00:49:32.500 I have other ones that I could share with you, so get in touch and we'll see how close we can match you up. 00:49:36.330 --> 00:49:36.780 Great. 00:49:36.790 --> 00:49:38.340 And I wanted to thank. 00:49:38.350 --> 00:49:44.090 I noticed Dan had shared Montana's call key in the chat and also. 00:49:44.790 --> 00:50:00.640 And we had shared in the chat that guide to processing and just so so folks are aware with a lot of guidance I've noticed across the net the network both through NABat and and other partners. 00:50:01.270 --> 00:50:12.100 We're kind of reaching the five year mark with a lot of this guidance and so you'll see a lot of stuff being updated and a across the network. 00:50:12.110 --> 00:50:20.490 So stay tuned to that are are there any other questions out there or we can go ahead and get started with our next speaker? 00:50:25.930 --> 00:50:26.250 OK. 00:50:29.160 --> 00:50:31.220 So we're going to jump right in. 00:50:31.740 --> 00:50:41.050 And Sarah from the NA back coordinating team, is going to to share and pass over to Frankie. 00:50:41.060 --> 00:50:42.290 Also on the team. 00:50:42.640 --> 00:50:43.900 So, Sarah, when you're ready. 00:50:46.270 --> 00:50:47.310 I think Frankie's doing both. 00:50:47.700 --> 00:50:50.080 OK, my apologies. 00:50:50.180 --> 00:50:50.780 So, Frankie. 00:50:56.470 --> 00:50:59.360 Sorry, could you could I go right after the next one? 00:51:00.820 --> 00:51:01.320 OK, OK. 00:51:02.120 --> 00:51:03.080 Does that mess anybody up? 00:51:04.220 --> 00:51:04.730 OK. 00:51:04.740 --> 00:51:08.600 So, Jason, Are you ready or would you be available to jump in? 00:51:11.100 --> 00:51:11.710 Yeah, sure. 00:51:11.720 --> 00:51:12.890 Just give me one quick second here. 00:51:13.470 --> 00:51:16.830 OK, luckily we have a little extra time so. 00:51:17.520 --> 00:51:17.880 Thank you. 00:51:20.890 --> 00:51:22.430 Sorry about this, OK. 00:51:24.060 --> 00:51:24.670 Thank you, Jason. 00:51:25.400 --> 00:51:25.630 Yeah. 00:51:25.640 --> 00:51:26.000 No worries. 00:51:26.010 --> 00:51:27.900 Uh, so you should be able to see my presentation now. 00:51:29.100 --> 00:51:29.220 Yes. 00:51:29.270 --> 00:51:34.010 Many of you will remember that from the first day of this workshop. 00:51:35.800 --> 00:51:44.630 NBC we're using Python to assemble all metadata from all recordings into a single data table with GRTS and quadrant information already pulled out. 00:51:44.640 --> 00:51:45.200 There automatically. 00:51:46.380 --> 00:51:54.420 So we merge to this table with relevant site and field form information and covariates like ambient temperature and humidity within. 00:51:54.460 --> 00:51:58.940 We then run a number of data checks and merge in final and manual IDs into that same table. 00:51:59.610 --> 00:52:03.570 For those of you who are interested in the specifics of that step, I did go over that in more detail in the first day's talk. 00:52:07.380 --> 00:52:10.010 Now we upload this information to the USGS. 00:52:10.020 --> 00:52:17.120 We'll be using the full metadata bulk upload CSV template and I'll just pull my laser pointer here. 00:52:17.690 --> 00:52:23.740 Uh, the one you pull from the website from the upload server data and station acoustic point? 00:52:23.830 --> 00:52:28.190 It will be generating another data table with similar columns to that one too. 00:52:28.240 --> 00:52:39.110 Upload through that system so to start this we prepare a second copy of our recording metadata table, abbreviated to only the column headers that are specified by the USGS bulk upload form. 00:52:40.020 --> 00:52:46.960 Nearly all the columns in that upload form are already represented in our data table, so this is just a simple step of copying and renaming column headers. 00:52:48.350 --> 00:52:52.020 The script then prints out a number of missing or. 00:52:52.030 --> 00:52:57.540 Sorry, a list of any missing headers and columns that was unable to match and we manually address those before moving to the next step. 00:52:57.550 --> 00:53:02.920 Whether it's just we forgot to add in a specific column or it was labeled something other than the script expected. 00:53:05.240 --> 00:53:09.090 Next week, clean and Recode information to fit formats recognized by the USGS system. 00:53:10.070 --> 00:53:11.550 Just a couple notable changes here. 00:53:11.560 --> 00:53:16.890 We convert date times to year, month and day separated with a T from hours and minutes and seconds. 00:53:17.300 --> 00:53:29.800 I like this letter format because it it if we need to edit the CSV manually in Excel after printing it out from Python, excel won't automatically recognize and convert the date to its own numeric format and it keeps that as a string. 00:53:29.850 --> 00:53:36.100 So it's easy to save and keep keep in the same format, otherwise Excel would change it. 00:53:37.790 --> 00:53:48.240 We split the species set out from the rest of the text in the class Scope, Pro classifier settings, metadata column and match that to a paired table containing the names of species sets we previously entered into the USGS system. 00:53:48.590 --> 00:54:07.640 So instead of being a list of five to 15 species, you see here like Codo app Foo, Yuma and so on and so forth, it gets summarized as the simple string name of BCC set SE1 and then otherwise we we we code the colonoscope classifier column text to indicate the version of classifiers in the format recognized by the USGS system. 00:54:07.830 --> 00:54:12.120 So instead of bats of North America 5.4, it changed it over to client scope 5.4 point X. 00:54:14.640 --> 00:54:20.950 Some of our recordings also have met multiple manual labels which were applied when activity from multiple bats were captured in a single recording. 00:54:21.840 --> 00:54:29.710 This script automatically splits the manual label column and uses the pandas explode function to create a duplicate row for each successive ID in the same recording. 00:54:31.350 --> 00:54:40.610 Values in the software type and auto ID columns that of that new row are change to no auto ID and a blank cell for the rows that contain additional manual IDs automatically. 00:54:42.880 --> 00:54:48.920 And similarly we create additional rows for the second auto ID label as well as last step. 00:54:48.930 --> 00:54:52.620 We adjust the manual labels to fit the sets required by the USGS system. 00:54:53.230 --> 00:55:00.200 For example, during our manual vetting, we use a lower case M prefix before all of our labels to denote that they were applied during manual analysis. 00:55:00.210 --> 00:55:04.650 But we stripped this prefix from the IDs and the manual ID column of the bulk upload form. 00:55:06.990 --> 00:55:10.540 And with that the script prints out of finalized form and we export. 00:55:10.590 --> 00:55:13.410 We upload that to the USGS bulk upload system. 00:55:14.640 --> 00:55:19.430 They're occasionally fears that escape the automated data checks, and we need to go back and rerun and and address. 00:55:19.440 --> 00:55:25.720 But having the data table printed out automatically in the correct format and they makes each step quite a bit lower effort. 00:55:28.160 --> 00:55:29.320 And yeah, thanks for your time. 00:55:31.850 --> 00:55:32.250 Great. 00:55:32.260 --> 00:55:35.540 Thanks so much and thanks for adapting on the fly, Jason. 00:55:35.550 --> 00:55:38.750 So Jason's shared his Python script for. 00:55:39.930 --> 00:55:43.830 Transferring results to NABat and so Frankie, are you available? 00:55:47.970 --> 00:55:48.910 OK, great. 00:55:48.530 --> 00:55:49.300 Yeah. Yes. 00:55:49.310 --> 00:55:49.560 Yeah. 00:55:49.570 --> 00:55:50.120 Thank you for that. 00:55:50.130 --> 00:55:50.640 So your minutes? 00:55:52.250 --> 00:55:54.660 Umm, alright, so let me share my screen real quick. 00:55:59.140 --> 00:55:59.380 1. 00:56:07.130 --> 00:56:10.310 That OK is that my presenter view is that the regular view? 00:56:12.010 --> 00:56:15.930 I can see I can see it without your prisoner notes, yeah. 00:56:15.890 --> 00:56:16.280 Awesome. 00:56:16.510 --> 00:56:17.180 OK, cool. 00:56:17.190 --> 00:56:26.130 So I'm gonna talk about a little bit today how what the some tips I've found for dealing with uploading a lot of audio files. 00:56:26.970 --> 00:56:46.210 Umm, I don't personally manage a ton of audio files for people, but as a technical outreach layers on with NABat, I do frequently get users coming to me with different questions or issues that they're running into with their Internet process. 00:56:46.220 --> 00:56:54.940 So I've had some users come to me saying that, you know, they had a bunch of audio files to upload to the Nabat partner portal. 00:56:55.250 --> 00:57:08.790 But and just the process of, you know, compressing them down into those tartal files that make the process go smoother, it can be really tedious to have to click through everything manually. 00:57:09.260 --> 00:57:24.260 So before I show you what I've found out for that, I'll just point out that you know, uploading your acoustic files isn't a requirement for NABat it's it's not something that, umm, you should let keep you. 00:57:24.270 --> 00:57:30.800 Like if you can't upload them, don't let that keep you from sharing your other call meta data that you've collected via surveying. 00:57:30.810 --> 00:57:35.360 But it is great if you can supply those files as well. 00:57:35.790 --> 00:57:38.470 It you know, it's another place where they can live. 00:57:38.480 --> 00:57:39.740 It's a backup for you. 00:57:40.410 --> 00:57:44.380 It's something we can come back to and review in the future as technology progresses. 00:57:44.390 --> 00:57:53.060 If we want to do more standardized analysis with the across all the projects that have those files provided, so it is great to provide those. 00:57:53.340 --> 00:57:53.440 Yeah. 00:57:54.330 --> 00:58:00.020 Anyway, so 7 zip is the program that we will be using to do this. 00:58:00.630 --> 00:58:02.780 I'm not gonna show you how to download 7 zip. 00:58:02.790 --> 00:58:05.540 You'll just go in to you can. 00:58:05.870 --> 00:58:08.710 It's very simple to download that and open it up. 00:58:09.460 --> 00:58:15.150 Umm, So what we're doing basically is using umm the. 00:58:18.100 --> 00:58:23.100 Uh, Microsoft has batch files. 00:58:23.110 --> 00:58:38.200 If you're using the Windows system where a very simple line of code in a text file and let me, uh, it's like this over, does that also show up on the screen? 00:58:40.190 --> 00:58:40.350 Yes. 00:58:40.280 --> 00:58:40.900 Yes it does. 00:58:41.360 --> 00:58:42.090 OK, great. 00:58:42.450 --> 00:58:49.100 So you can see here the way I've organized this is the an example of like organizing my. 00:58:50.440 --> 00:58:51.330 Acoustic files. 00:58:51.340 --> 00:58:53.360 We have two folders here from two different sites. 00:58:54.420 --> 00:58:56.980 They each have acoustic files contained inside of them. 00:58:58.850 --> 00:59:15.020 And umm, what we've done here is we've saved 2 batch files that I'm gonna talk about here in just a second and they're not back, which is great for this have SO2 batch files and what these are doing, this is the first one. 00:59:15.030 --> 00:59:40.080 You'll see that to create this batch file, all you have to do is open up the folder that you're working in and hit go down to new and text document and all these are are text documents that you fill out and save under the dot bat file name file extension. 00:59:41.860 --> 00:59:47.340 So those when you are looking at them, I'll walk through the what that looks like real quick. 00:59:49.050 --> 00:59:49.350 Yeah. 00:59:49.360 --> 00:59:50.480 Sorry dude. 00:59:50.780 --> 00:59:51.190 Nope. 00:59:51.780 --> 00:59:52.100 I went. 00:59:52.110 --> 00:59:52.570 There we go. 00:59:53.220 --> 01:00:01.800 The first one that you'll run looks like this, so if you're wanting to do something like this, you could screenshot this or come back to this presentation later. 01:00:02.180 --> 01:00:07.750 This is all you have to put in that file and just to walk through the code really quickly. 01:00:08.020 --> 01:00:10.510 This first section initiates a loop. 01:00:11.160 --> 01:00:11.630 Umm. 01:00:12.160 --> 01:00:22.030 Via the Microsoft like operating system, when you put the file, it'll initiate this loop and it's gonna start working through the all the directories in your current directory. 01:00:22.360 --> 01:00:44.540 So this is gonna start umm, walking through the different folders that you have in the same directory as this batch file and then umm it's gonna go in and look at the very real name and assign that as the directory name that it's reading through this loop. 01:00:46.070 --> 01:00:58.970 It's gonna call upon 7 zip executable files, so all you have to do is have 7 zip installed on your machine and it's gonna call upon that and it's going to create a new archive. 01:00:59.360 --> 01:01:08.720 This is basically compressing the directories that it's working through, and it's telling it to create that archive in the tar format. 01:01:10.350 --> 01:01:20.070 And then it's just saying go through and fill that archive that it just created with your variable from early from before working through this. 01:01:20.500 --> 01:01:28.330 So to demonstrate that really quickly, you just again, here's your directories that you're working with. 01:01:29.350 --> 01:01:47.440 Uh, I'll just click this double click this and get popped up so quick on my other screen it just the command prompt opens for a brief moment and runs through the code and you can see that it read through these directories and it created tar files. 01:01:48.700 --> 01:01:49.180 Boring. 01:01:49.230 --> 01:02:07.360 So to compress those one step further into that dot tar dot GZ, umm that really compresses things down to put them up to our uploading process I'll just click this one and uh you can see that it does the same thing. 01:02:07.370 --> 01:02:09.330 It steps through umm. 01:02:09.770 --> 01:02:10.800 I'll show you in the PowerPoint. 01:02:13.610 --> 01:02:14.380 It's not where it is. 01:02:16.600 --> 01:02:24.060 This is the second set of code for that second batch file and like before, yes starting a loop. 01:02:24.490 --> 01:02:29.310 This time it's looking for each tar file in the directory that that batch file is sitting in. 01:02:30.280 --> 01:02:32.750 Uh, it's going to name the new thing. 01:02:32.760 --> 01:02:33.150 It's or. 01:02:33.160 --> 01:02:35.990 It's gonna read in the variable name as the file name that it's reading through. 01:02:37.030 --> 01:02:39.890 It will then call upon the seven zip executable file. 01:02:41.100 --> 01:02:45.310 It'll go in to create a new archive and it's gonna be naming it under this format. 01:02:45.320 --> 01:03:01.320 Now the dot tar dot GZ format umm in this little till the end just is removing the for the file format from before just to make the naming work and then it's going to fill that new archive with the files that it's working through in this loop. 01:03:02.330 --> 01:03:07.680 So again, you know, all you have to do is have these batch files. 01:03:09.230 --> 01:03:09.650 Uh. 01:03:10.130 --> 01:03:14.320 Save in your folder that you're working in your directory that you're working in. 01:03:14.570 --> 01:03:28.470 Click these and it'll run through pretty quickly that stepped through these different uh loops to call upon 7 zip and automate this process for you, and it's just something you can leave running in the background. 01:03:28.480 --> 01:03:44.330 If you do have a whole lot of folders and a whole lot of files, umm, and then obviously you can go in and delete what you don't need after the fact and then upload these dot tar dot GZ files to your partner portal project real quick. 01:03:46.000 --> 01:04:01.090 Uh, something to keep an eye out for is are upload our partner portal does have a 3 GB limit, so you'll wanna experiment just a tiny bit to make sure that when you're compressing those files down, your folders are subset. 01:04:01.100 --> 01:04:07.890 The data subsetted such that the final product comes out to 3 gigabytes roughly or below. 01:04:09.140 --> 01:04:09.470 Umm. 01:04:09.480 --> 01:04:19.160 And then also when you're going to upload these files that hurdle hurdle that still exists, that can be frustrating is that there's a 15 minute timeout by science base. 01:04:19.760 --> 01:04:20.220 Umm. 01:04:20.720 --> 01:04:22.210 As just like a security measure. 01:04:22.220 --> 01:04:23.550 So you'll want. 01:04:23.560 --> 01:04:38.240 You'll need to be active on NABat to keep active on NABat on the partner portal to make deep sure these uploads go through, but you can time those right when you're just kind of working throughout the day and it can just run in the background while you're doing that. 01:04:39.190 --> 01:04:39.940 And with that. 01:04:43.210 --> 01:04:43.960 That's all my stuff. 01:04:43.970 --> 01:04:45.330 This is my contact information. 01:04:45.340 --> 01:04:53.810 Uh, if you didn't screenshot those, or if you want any help with this, feel free to reach out to me anytime I'm also available on the Internet website. 01:04:55.950 --> 01:04:56.310 Great. 01:04:56.320 --> 01:04:57.900 Thank you so much, Frankie. 01:04:57.950 --> 01:05:00.790 So now we're gonna pass it along to Jeff. 01:05:02.050 --> 01:05:03.060 With CMI. 01:05:03.110 --> 01:05:04.250 So when you're ready, Jeff? 01:05:08.670 --> 01:05:09.060 Hey there. 01:05:09.030 --> 01:05:11.880 All right, you show my screen here. 01:05:19.240 --> 01:05:20.440 OK, we've got it, Jeff. 01:05:21.320 --> 01:05:21.870 Got it. 01:05:21.880 --> 01:05:22.430 Great. 01:05:22.620 --> 01:05:31.430 OK, so I'm Jeff Schluter with conservation metrics and we partner with Bat Conservation International to make up the national data processing lab. 01:05:31.440 --> 01:05:35.440 We work with NABat data from a bunch of different collaborators all over the country. 01:05:37.990 --> 01:05:50.680 As my coworker is carrying Jack described yesterday, we have a very complex pipeline for processing NABat data of all types and at the end of that pipeline we have two types of data. 01:05:50.730 --> 01:05:59.630 We have vetted spreadsheets and the vetting is done by Kristen at BCI and we have the sound files that have been processed and attributed. 01:06:01.310 --> 01:06:07.500 Uh, we process a huge amount of data each year, millions of files and usually more than 10 terabytes. 01:06:07.850 --> 01:06:18.990 And because of that we transfer the data to USGS through Amazon Web Services rather than using the any of that website like Frankie just described. 01:06:20.240 --> 01:06:26.610 There's three components to transferring data this way, reorganizing it, uploading it and sharing it. 01:06:26.620 --> 01:06:29.090 And I'll kind of briefly go through that workflow. 01:06:31.110 --> 01:06:37.890 So first we reorganize so we process data locally on our servers and computers at our office. 01:06:38.480 --> 01:06:43.940 Uh, and they're stored in a file structure that makes sense for the way that we process the data. 01:06:44.010 --> 01:06:52.760 So underneath the data ID which is our unit of data processing, we have subfolders for each grid cell and site. 01:06:55.340 --> 01:06:58.830 And we need to convert that to the. 01:07:00.140 --> 01:07:21.650 NABat file structure in order to transfer the data to AWS and that just has they have subfolders for each project ID which can be found in the URL for the the project in the NABat website and so we need to kind of crosswalk our data structure to to theirs before we can do the transfer. 01:07:22.710 --> 01:07:32.510 We have a shared Google sheet that we collaborate on with BCCI and we can join the information about the files that we have locally. 01:07:33.190 --> 01:07:43.520 We can join that to the information from this Google sheet to attach the project ID values for each data ID and create new directory paths for each file. 01:07:45.200 --> 01:07:45.750 Umm. 01:07:46.940 --> 01:07:48.650 And then we can start the transfer. 01:07:48.660 --> 01:08:11.160 So first step for reorganization is we copy our files to the new AWS directory structure on our local storage and there's an example of kind of how the files are saved in our system versus the path that is needed for getting the data into USGS that system. 01:08:13.670 --> 01:08:22.070 After that, we can uh create a new empty bucket in our AWS account. 01:08:22.080 --> 01:08:26.980 So we have a CMI, we store a bunch of data and AWS for backup and things like that. 01:08:26.990 --> 01:08:32.390 And a bucket is Aws's term for data storage unit. 01:08:33.130 --> 01:08:35.230 So we create an empty bucket in our account. 01:08:37.470 --> 01:08:42.100 Next weekend, sync our local directory to that new bucket. 01:08:42.570 --> 01:08:55.150 EWS has a command line interface AWS CLI which is a series of functions that can run on a local machine and interact with Amazon's Web storage. 01:08:55.160 --> 01:08:56.420 's web storage. 01:08:57.190 --> 01:09:02.060 There's sync function is really helpful. 01:09:02.070 --> 01:09:07.210 Function has great multithreading so it can transfer file multiple files simultaneously. 01:09:08.150 --> 01:09:17.320 Umm it is the fastest way that we've found to upload data to AWS and it's robust for like timeouts or lost connections. 01:09:17.330 --> 01:09:31.710 It checks the completeness of files and it is a great sync function, so we transferring data to the cloud can be fraught sometimes with timeout or lots of different issues, and this is a great option for doing it successfully. 01:09:32.460 --> 01:09:46.390 Umm, that software that Amazon Command Line interface runs on a local machines terminal and we can pass commands to it through through R which is kind of our language of choice for manipulating data. 01:09:48.570 --> 01:10:06.080 Umm, the next step we can then delete the local copy of the data that it's in AWS structure, because we've already transferred it now to a WS and it would be nice to not make duplicate copies of all of our data on our local storage. 01:10:06.150 --> 01:10:20.610 It takes time to make those copies, and it takes a lot of room when we're working with big data sets, but that is the easiest way that that we found to use the sync function because the transfer of the data transfer is really the time intensive part of the process. 01:10:22.530 --> 01:10:37.280 Umm, now that the data are in the cloud, we need to share our AWS bucket with USGS and you can control all types of parameters about your buckets and using bucket permissions. 01:10:38.710 --> 01:10:53.160 Umm, there's a bucket configuration and permissions are controlled by a Jason file, so it's like a special text file, a Jason format and you can upload this text file. 01:10:53.220 --> 01:11:04.530 Well umm, in the AWS web interface you just basically paste the Jason format into a text block and that changes the permissions for the bucket. 01:11:05.240 --> 01:11:11.500 Umm and this example that I stuck up here gives permission for three different types of operations. 01:11:11.510 --> 01:11:17.000 So list bucket get object and get object tagging which are kind of the three permissions that. 01:11:18.260 --> 01:11:27.270 Collaborator needs in order to see what you have in your bucket and pull it across to their account and it gives the access to two different account IDs. 01:11:27.280 --> 01:11:33.430 We have two different accounts that we were given by USGS Umm. 01:11:33.780 --> 01:11:51.270 The next step I send an email and USGS get takes the time to copy the data over to their own account and once they say that it's done, we can delete our bucket and we often I got suit. 01:11:51.410 --> 01:11:51.970 Good to know. 01:11:51.980 --> 01:11:55.000 We often just wait till the end of the year or the end of the season. 01:11:55.010 --> 01:11:56.170 We have a whole bunch of data. 01:11:56.180 --> 01:11:57.780 It's all been processed uniformly. 01:11:58.730 --> 01:12:08.280 All the the data sheets have been uploaded to NABat already and then this is just like a bulk data transfer at the end of the whole analysis process. 01:12:10.710 --> 01:12:13.260 Once that's done, we can delete our bucket and that's that. 01:12:13.270 --> 01:12:14.090 Everything's transferred. 01:12:15.580 --> 01:12:16.160 Thanks. 01:12:17.370 --> 01:12:17.760 Great. 01:12:17.770 --> 01:12:23.570 Thanks so much Jeff for sharing your workflow with using Amazon Web Services. 01:12:23.580 --> 01:12:27.590 So now we're going to pass it along to Patrick. 01:12:29.290 --> 01:12:32.850 He's going to share some R scripts and additional checks that they have. 01:12:36.340 --> 01:12:39.050 OK, everybody, I'm Patrick emblidge. 01:12:39.100 --> 01:12:48.830 I am the data manager for the Northwest Bat Hub, and I'm gonna review our process for preparing stationary acoustic data and uploading it to NABat. 01:12:49.080 --> 01:13:15.190 It's gonna be pretty similar to JSON's, but we use our instead of Python, and for those of you who are with us for Tuesday session, this process might feel like it has a lot of similarities with the some of the QA QC talks, but I think that's OK because it's important to constantly be on the lookout for unexpected data errors and you never know when a novel error is gonna evade your earlier checks. 01:13:16.420 --> 01:13:19.970 So I like to just do as many error checks as possible. 01:13:21.100 --> 01:13:25.140 I'll start with a little schematic to show the workflow of the process we use. 01:13:26.950 --> 01:13:57.670 We format the metadata and acoustic output from vetting separately, but it's following similar processes and then we join them to upload to NABat and one of the first steps for preparing metadata for NABat upload is to combine all sources of metadata and for us we have metadata that was collected in field maps and metadata from paper data sheets in separate Excel files. 01:13:58.220 --> 01:13:58.670 Umm. 01:13:59.320 --> 01:14:08.610 And as a general rule, when you need to combine data sets, you need to make sure that umm field names are consistent. 01:14:08.680 --> 01:14:14.340 Field data types are consistent and units are consistent across data sets. 01:14:15.850 --> 01:14:16.130 Umm. 01:14:16.470 --> 01:14:26.750 Field names and data types are pretty straightforward, so I'll give an example of how we deal with how we address different units between datasets. 01:14:27.580 --> 01:14:32.480 Umm in field maps date time data types? 01:14:32.910 --> 01:14:39.690 Umm are stored in UTC, while on paper data sheets there's stored as the local time. 01:14:40.460 --> 01:14:50.210 Umm, so we need to convert UTC to local time and store it as a text type as Jason said. 01:14:50.220 --> 01:14:58.060 So it excel doesn't adjust it and also because our does not tolerate different time zones in the same column. 01:14:59.960 --> 01:15:22.220 Uh, so first we need to identify the time zone of each station location based on its geographical coordinates, and then we use the time zone to reformat time from UTC to the local time and store it as a character object. 01:15:22.230 --> 01:15:26.070 So then we're ready to combine the two datasets. 01:15:28.650 --> 01:15:35.870 And next a later step will be joining the acoustic data with the metadata. 01:15:35.880 --> 01:15:44.010 So we need a unique identifier for each deployment and in Tuesday's session you saw different people did it different ways. 01:15:44.360 --> 01:16:01.010 But we use the combination of the grid cell ID, the location name and a site deployment number, which is basically just a sequential number for sites that were surveyed more than once throughout the season. 01:16:01.020 --> 01:16:30.190 So we first uh, just display all station locations that were surveyed more than once, and then we use those that output to define uh site deployment numbers for each deployment, and the last step for metadata is to format the fields to match the Nabat format, and they're accepted values. 01:16:31.890 --> 01:16:39.190 I also use this opportunity as another quality control step and I'll go through each each field 1 by 1. 01:16:39.990 --> 01:16:40.460 Umm. 01:16:40.650 --> 01:16:47.400 View the array of values that are contained within the data set to make sure there's nothing obviously erroneous. 01:16:48.010 --> 01:16:52.190 As one example for microphone height. I have it. 01:16:52.200 --> 01:17:01.370 Display any values that are greater than 4 meters, which is usually caused by an observer using feet instead of meters as their unit of measure. 01:17:01.910 --> 01:17:12.540 And I'll I can confirm by looking at site photos or asking the surveyor directly and then making corrections into the metadata as needed. 01:17:13.440 --> 01:17:14.500 Umm after that. 01:17:17.450 --> 01:17:25.740 We then perform a similar process with the acoustic data where we compile all the betted outputs into a single data set. 01:17:26.090 --> 01:17:31.680 We assign unique identifiers with the site deployment number and we format the fields for NABat. 01:17:35.220 --> 01:17:37.900 Next, we join the metadata and the acoustic data. 01:17:38.790 --> 01:17:45.010 Umm, according to the Grits location, name and site deployment number. 01:17:45.970 --> 01:17:52.240 Uh we do a we use a full join so that we retain records for all deployments. 01:17:52.250 --> 01:18:22.360 Whether or not processed acoustic results are present and some situations where they wouldn't be, as if there's a detector failure for whatever reason or a poor placement so that all files are scrubbed out as noise, I still want to document our effort, so I'll add notes for those types of deployments in the unusual occurrences field to to identify these errors. 01:18:24.430 --> 01:18:24.770 Umm. 01:18:24.910 --> 01:18:42.320 And as one final check, we identify the acoustic files that occur outside of the deployment time and review them to determine if the files should actually be removed or if the start and stop times were added or were entered incorrectly. 01:18:42.840 --> 01:18:49.370 And it might seem late to be doing this, but since this is an automated process, it's really easy to just go back and run this whole script again. 01:18:50.550 --> 01:18:57.360 Once that's done, we add it, we append our output to the NABat template and it's ready to upload. 01:18:57.680 --> 01:19:11.090 So we can either manually upload it through the Nabat partner portal or we can upload it directly through our using the Nabat R package and that leads us into the questions portion of the session. 01:19:13.550 --> 01:19:13.900 Great. 01:19:13.910 --> 01:19:19.060 Thanks so much Patrick for sharing your script and and approach. 01:19:19.710 --> 01:19:23.560 So we have just a couple minutes for questions. 01:19:23.570 --> 01:19:28.980 Let's focus on if we can right now on transferring results to the partner portal. 01:19:29.770 --> 01:19:36.920 There were a couple of questions really that popped up in the chat related to the Amazon Web Services. 01:19:37.390 --> 01:20:02.260 And so Ben niece was curious if other partner portal or partner projects and had access to a WS and Bethenny reiterated that yes, that that is a possibility for all projects as long as you've fulfilled the need of having those Amazon Web service assets that are required. 01:20:02.930 --> 01:20:03.380 Umm. 01:20:03.650 --> 01:20:13.890 And then there were also some questions related to the amount of time associated with uploading through AWS. 01:20:14.390 --> 01:20:21.790 So if Jeff, if you wouldn't mind popping back on and just briefly sharing some details related to that? 01:20:25.070 --> 01:20:28.510 As well as compression, I saw that was also popping up in the chat. 01:20:29.660 --> 01:20:30.070 Yeah. 01:20:30.080 --> 01:20:34.090 So we just transfer the the raw files across. 01:20:34.100 --> 01:20:39.320 We don't do touring by night or by sight or anything like that. 01:20:39.480 --> 01:20:44.480 My understanding is that the so the sub the project ID. 01:20:45.980 --> 01:20:51.870 There's one folder per project ID, and then underneath that are just all the files and the file names. 01:20:52.540 --> 01:21:03.530 Those are the attributed files and the file names correspond to entries in the the bending spreadsheets and the you know the metadata sheets that were that were uploaded. 01:21:03.600 --> 01:21:03.840 Good. 01:21:09.020 --> 01:21:17.510 And so you were also mentioning that to transfer one terabyte of data across the web services takes approximately an hour. 01:21:19.140 --> 01:21:20.170 Takes about an hour. 01:21:20.180 --> 01:21:20.490 Yeah. 01:21:20.500 --> 01:21:28.470 So we have a really fast Internet connection and our office, but it's all about the the volume of data and how fast your connection is. 01:21:29.020 --> 01:21:30.010 So it'll vary. 01:21:30.060 --> 01:21:49.290 It varies for us depending on the time of day and how much web traffic there is, but it's a process that just can run in the background and they all start at like at the end of the day when we're getting ready to leave work and let it run overnight and it's done in the morning when I come back. 01:21:54.220 --> 01:21:54.580 Great. 01:21:54.590 --> 01:21:55.450 Thank you, Jeff. 01:21:55.500 --> 01:21:59.870 And did we have any any other questions from those out in the audience? 01:22:10.490 --> 01:22:11.980 Yeah, I I guess I'll just share too. 01:22:11.990 --> 01:22:17.190 I know I was kind of breezing over AWS and Amazon Portal. 01:22:17.200 --> 01:22:30.950 If you haven't worked with AWS that probably wasn't necessarily intuitive but so Amazon has cloud services so it's you can use cloud computing and cloud storage. 01:22:30.960 --> 01:22:59.160 They call it S3 storage and it's pretty straightforward to set up an account, and there's lots of tutorials online for how to have an S3 account and how to create a bucket, and with a little bit of like setting up the CLI, the command line interface tool on your local machine took a little bit of finagling, but once you get that set up, the tools work really smoothly. 01:23:00.180 --> 01:23:08.620 Like I said, there's a lot of documentation, so if you know you have some tech savvy, it's not a huge barrier to get an account set U and create buckets. 01:23:13.850 --> 01:23:14.220 Great. 01:23:14.230 --> 01:23:15.140 Thank you, Jeff. 01:23:15.190 --> 01:23:16.900 So, umm what? 01:23:16.910 --> 01:23:31.160 We'll go ahead and do is we'll move into our portion of the workshop focused on automating cooperators reports and different approaches folks have across the network with that. 01:23:31.270 --> 01:23:34.820 And then we'll have more time to kind of dive deeper into. 01:23:36.780 --> 01:23:46.450 Questions that have come up throughout the workshop today, but also if there were any lingering questions from session one, that'll be a great forum for that. 01:23:46.460 --> 01:23:52.730 So now Kathy and Christian, if you're available. 01:23:56.040 --> 01:23:56.590 Hi. 01:23:56.680 --> 01:24:06.310 Yeah, I'm gonna be presenting on behalf of BCCI and our reporting, so give me a second to share my screen. 01:24:16.140 --> 01:24:16.530 Alright. 01:24:16.540 --> 01:24:16.950 Do you see this? 01:24:21.060 --> 01:24:22.210 Alright, sounds good. 01:24:22.370 --> 01:24:23.080 Yes, sorry. 01:24:22.320 --> 01:24:25.410 It's a thumbs up from your thanks. 01:24:26.270 --> 01:24:26.870 OK. 01:24:26.880 --> 01:24:27.810 Hi everyone. 01:24:28.020 --> 01:24:29.950 My name is Kathy Gerst. 01:24:29.960 --> 01:24:41.540 I am a conservation research coordinator at BAT Conservation International and I also am the Southwest Nabat Hub coordinator for Arizona and New Mexico. 01:24:43.190 --> 01:25:17.620 And I'm presenting here on behalf of sort of the reporting arm of the national Data processing lab, the collaboration you've heard about a few times, most recently from Jeff just now about between conservation metrics and bat Conservation International and the workflow that I'm gonna show you for creating a customized results report for any about partners was created by our former PacWest Hub coordinator, Nat Goodbee, as well as with Kristen Long who's on this call, who's our acoustic specialist. 01:25:17.670 --> 01:25:32.970 And she also is the one that's generating these reports and for our partners and will be available for more detailed walkthrough of the R markdown files that I'm going to show you if you're interested in all this, will be available for sharing. 01:25:34.890 --> 01:25:36.050 OK, so. 01:25:38.050 --> 01:25:38.720 Ohm. 01:25:39.190 --> 01:25:47.590 I'm gonna show you here examples of the two types of reports that we create for our partners right here. 01:25:47.600 --> 01:25:52.340 We it says landowner report inputs and here it says NDP L report inputs. 01:25:52.760 --> 01:26:11.070 And so the landowner reports, generate a more simple a simple output and give a more concise output for from surveys, whereas the detailed NDP L partner reports that agency partners and biologists receive so. 01:26:11.080 --> 01:26:21.030 These are the reports that we generate for Forest Service or conservation organizations, Park Service, state wildlife agencies, etcetera. 01:26:21.860 --> 01:26:41.420 Every time these run, we're tapping into both the Nabat API as well as the Google spreadsheet I all that we share with conservation metrics to manage all the information about our data set ID and and sort of you know the species, the species that are used for the processing. 01:26:42.810 --> 01:26:50.220 And you know, I we decided not to sort of go through a detailed rundown of everything in the R markdown. 01:26:50.230 --> 01:26:55.580 But like I said, we're willing to share these, but like for here, this is our land owner report input. 01:26:55.790 --> 01:26:59.820 We I can give very simple information. 01:26:59.830 --> 01:27:02.640 Just the GRTS ID number. 01:27:02.910 --> 01:27:10.040 We can also produce reports either for all the deployment locations within a cell or just one. 01:27:10.050 --> 01:27:16.550 So I'm gonna show you an example for just one location called Wow, Arizona. 01:27:17.110 --> 01:27:29.420 We tap into the project ID and we also customize these depending on if it's for the Southwest hub or the Pacific or sorry, the PAC W Hub for the more detailed reports. 01:27:29.430 --> 01:27:34.720 We also can do mobile transect reporting and give a little bit more context. 01:27:36.080 --> 01:27:38.090 About the agencies etcetera. 01:27:40.570 --> 01:27:43.910 So I am going to now just. 01:27:48.970 --> 01:27:53.000 Umm, so now I'm going to show you 2 examples of these two types of reports. 01:27:53.210 --> 01:27:55.480 1st I'll show you a landowner report. 01:27:55.930 --> 01:28:00.840 These are sort of the highlights of what we offer that's different about these reports. 01:28:00.850 --> 01:28:02.840 These are for a landowner audience. 01:28:03.230 --> 01:28:15.300 The private land owners these tend to go towards private land owners that gave us permission to do surveys on their land, but they are not necessarily doing the surveys themselves or need this information for management purposes. 01:28:15.310 --> 01:28:19.130 So we make them a bit more accessible for sort of a general audience. 01:28:20.610 --> 01:28:25.030 The they can be made for one cell or a deployment location, and there's sort of. 01:28:25.620 --> 01:28:31.040 Uh, and emphasis on the species detected and photos of those species. 01:28:32.660 --> 01:28:41.890 So I'm going to pull up, umm, one of those reports, just to show you this is a report that we created for a landowner. 01:28:42.310 --> 01:28:50.370 Umm, that gave us permission to do surveys on at one location within a cell. 01:28:50.820 --> 01:28:55.460 And so we produced these HTML reports that are interactive. 01:28:55.470 --> 01:29:00.730 So on the right corner, we can Scroll down to different components of them. 01:29:02.270 --> 01:29:08.360 We start out the report just saying how mNABat species were detected in the survey locations. 01:29:08.830 --> 01:29:16.580 We produced these maps that will show both the location and the cell boundaries, and if you scroll over them you get the details. 01:29:16.590 --> 01:29:19.750 So these could be for multiple locations or just one. 01:29:21.240 --> 01:29:22.010 Umm. 01:29:22.070 --> 01:29:35.760 And then if we go here to species detected, we don't use the scientific names, we just use the the common names and just provide the list. 01:29:35.770 --> 01:29:44.870 And then if the location had been surveyed in other years, we could show we would have columns showing which species were detected and which years. 01:29:45.730 --> 01:29:56.620 So if we go down to photos of species detected and I open this up, we give and see the reports, generate the photos of all the species detected. 01:29:57.050 --> 01:29:58.620 These can be closed. 01:29:58.950 --> 01:30:05.100 They also show the stationary site photos, so these are pulled from the from the. 01:30:05.210 --> 01:30:15.340 NABat database and a species count graph that's useful if there's multiple locations or multiple years. 01:30:16.210 --> 01:30:31.270 We also give information on survey dates and bad activity in the format of the count of bat calls that were recorded throughout the night and then we save the information sort of on the program with background information and sort of threats to that. 01:30:31.280 --> 01:30:42.240 So at the bottom, along with methods and examples of call sonograms and then finish it off with some some links there. 01:30:43.710 --> 01:30:52.490 So if I go now, I'll show you an example of our more detailed reporting that go to these wildlife agency partners. 01:30:53.010 --> 01:31:05.770 The audience is again more towards the biologist and the agencies, and these get produced at the scale of anywhere from a location to a cell, but sort of at that park, forest or regional level. 01:31:05.940 --> 01:31:12.190 And there's more customization and context of the Nabat program and the details of the surveys that were done. 01:31:12.270 --> 01:31:12.410 Yeah. 01:31:13.460 --> 01:31:15.510 So I'll show you what that looks like. 01:31:15.520 --> 01:31:31.060 Now here's a report that was done for a National Forest and these reports, as you can see, are more heavy on the methods at the beginning of the report. 01:31:31.070 --> 01:31:36.610 But we still have sort of the same format here where we have an interactive HTML report. 01:31:36.950 --> 01:31:38.100 We have our methods. 01:31:38.970 --> 01:31:39.490 Same thing. 01:31:39.500 --> 01:31:52.000 We have a map of sea scenery locations, and this location also or sorry, this project also was doing mobile transect, so we can combine stationary and mobile transects in the same report. 01:31:52.010 --> 01:32:00.480 We can, umm, zoom in on on the cells, get information if we're hovering over on the cell ID in the survey. 01:32:00.650 --> 01:32:01.260 I don't. 01:32:01.310 --> 01:32:12.380 Sorry, the survey site names can give them a stationary photos that they sent in with their survey and metadata information. 01:32:12.710 --> 01:32:16.900 And then here we can see the map of the mobile transects as well, which can be really handy. 01:32:18.630 --> 01:32:44.620 What those look like, we also provide the potential species list that may occurred and what were used for to generate the reports and then here no part six, we give the stationary acoustic results in both sort of list format and as tables showing what was detected in each year that that was surveyed for all the different cells. 01:32:45.510 --> 01:32:47.760 So so, Kathy, I'm gonna. 01:32:47.850 --> 01:32:48.080 I'm. 01:32:47.430 --> 01:32:50.770 Yeah, you're gonna cut me off, OK? 01:32:48.120 --> 01:32:56.360 I'm gonna enter, interject and so these are really slick reports, but we're gonna pass it over to Jason. 01:32:57.240 --> 01:32:57.540 OK. 01:32:58.300 --> 01:33:00.170 Thank you so much for sharing. 01:32:59.980 --> 01:33:00.990 Yeah, no problem. 01:33:01.000 --> 01:33:05.500 Sorry, I went over but I I was just about to wrap it up anyways. 01:33:06.210 --> 01:33:07.830 OK, awesome. 01:33:07.840 --> 01:33:08.520 Thanks. 01:33:09.130 --> 01:33:14.550 And then Jason, if you're available to show us from the BBC's perspective? 01:33:16.010 --> 01:33:16.460 Sounds good. 01:33:16.470 --> 01:33:16.680 Yeah. 01:33:16.690 --> 01:33:21.520 And Kathy, a lot of the stuff that I'm gonna be doing is very similar to what you presented. 01:33:21.530 --> 01:33:25.860 So I can highlight some of the parts that you didn't have time to finish off hopefully. 01:33:27.470 --> 01:33:35.190 So in BC, we're also using the our markdown method to print out reports we send to our partners and landowners. 01:33:36.030 --> 01:33:41.620 We have a little bit more of a text heavy approach to it and I'm just sharing my PowerPoint now. 01:33:48.090 --> 01:33:57.570 So we sent these around in a PDF format as opposed to the HTML and we we do include a little bit more text summaries of the species and stuff like that, hopefully talk about them. 01:34:00.730 --> 01:34:11.280 You, you may note that I've been talking about Python for the first half of the process, but once those data tables are formatted and finalized, we switch over to R for the statistics statistical analysis and to produce these final products. 01:34:12.810 --> 01:34:23.510 So this is an our markdown report, I mean the introduction and the first half of this method section are basic text repeated to all grid cells within the second half of the methods. 01:34:23.580 --> 01:34:41.510 That latter paragraph you see here, we print out some basic information about the grid cell, like the first year that the grid cell was established, the region, the grid cell occupied is within, it falls within the total number of detectors deployed in that grid cell ever, and have that types of deployment locations, things like that. 01:34:45.020 --> 01:35:03.610 Next in the results section, we auto generate some text with fill in the blank information from the R script, filling in basic details about the current year, like total number of at recordings, number of stationary detectors deployed, number of transit replicates first and last date of recording, and percentage of recordings identifiable to species. 01:35:05.070 --> 01:35:08.290 We have a little bit of text discussing why these are interesting metrics and things like that as well. 01:35:09.290 --> 01:35:19.750 We also have a couple recording quality sentence sentences here that are pulled from a paired file that we enter comments to during our manual analysis because those would be a little bit too hard to actually automatically print out. 01:35:20.480 --> 01:35:36.220 So it just pulls it from that period file and the next paragraph talks about the species detected, total number, common names, any species that were detected in this grid cell that weren't in the past years that didn't show up this year in in new species detected this year that hadn't been recorded in previous years. 01:35:36.530 --> 01:35:42.840 And any species we expect may occur in this area based on this, species ranges that haven't yet been identified in the recordings. 01:35:43.030 --> 01:35:48.130 So the sentences are all automatically matched up versus those tables and printed out here in this paragraph. 01:35:51.370 --> 01:35:56.260 Our first table is a quick print out of annual species detections summarized by year end quadrant. 01:35:56.610 --> 01:36:01.980 We only report species presence and absence to avoid misleading conclusions. 01:36:01.990 --> 01:36:13.860 If the reader starts to compare between mean number of recordings between species within a year in a site, things like that, we also report obviously Knights and quadrant and mean nightly recordings. 01:36:14.990 --> 01:36:16.230 But for all species together. 01:36:18.070 --> 01:36:24.800 Species activity is a section that was highly requested by our partners, often asking what the most frequently recorded species were at their sites. 01:36:24.850 --> 01:36:30.740 So here we list the four most four species that show up in the most recordings at each site. 01:36:31.780 --> 01:36:32.790 Sorry at and. 01:36:32.800 --> 01:36:35.590 They're at each detector in their grid cell. 01:36:37.090 --> 01:37:00.290 We do caution that this doesn't necessarily mean those species at the most abundant, because it is kind of a mix of the species likelihood of being active near the surveillance detector sites, the distance each species pulses will travel through the air before their two fainted, detect the likelihood each species can be identified versus other nearby species, and the influence of covariates, as well as also the obvious actual differences in abundance. 01:37:01.320 --> 01:37:10.070 So with that, we actually match in a figure created in Arcmap through the use of data driven pages. 01:37:10.080 --> 01:37:17.500 This is just generated prior to the reports and it pulls the images by file name and matches them in and. 01:37:17.580 --> 01:37:24.540 You'll note I've kind of skipped figure one because that's just a basic number basic image of the mapping map locations of all grid cells and BC. 01:37:26.820 --> 01:37:46.020 For good cells that have more than five years of observation, we also fit a negative binomial GLM to each year's nightly activity using temperature, humidity, wind speed and degree days as covariates, just to give a simple very simple representation of how the activity in their grid cells is trending over the years of years of observation. 01:37:46.030 --> 01:37:50.450 After five years, some of our partners are to say, OK, so what's happening now? 01:37:50.460 --> 01:37:51.730 What's different kind of thing? 01:37:54.830 --> 01:38:04.690 Discussion once again starts to pull some more sentences from that paired file that I talked about those populated during a manual vetting for both the technical assessment and the rare or unexpected species paragraph. 01:38:04.750 --> 01:38:13.020 Basically we just add some sentences in manually that say this space is one that was unexpected in your grid cell, but we found it here in this year's analysis. 01:38:13.030 --> 01:38:21.380 Or perhaps something along the lines of recording quality decreased this year because X may have happened or something like that. 01:38:21.460 --> 01:38:23.240 Mechanical issues with the microphone, for instance. 01:38:25.580 --> 01:38:35.270 The rest of the discussion prints out specific paragraphs and sentences to provide some interesting characteristics about the species detected and comment on some of the habitat types used by the species found in the grid cell. 01:38:35.340 --> 01:38:43.770 For instance, if little brown, myotis and numerous are detected, the script will print out a paragraph talking about the OR. 01:38:43.780 --> 01:38:57.390 Sorry if both of those pieces are detected and the at least one of the detectors is placed near wetland forging habitat, this script will insert a paragraph talking about how both of those species NBC will forge on the adult stages of aquatic insects. 01:38:57.400 --> 01:39:13.440 So maybe seeing some activity because of that and the last step that we have here is to manually check over the discussion and the rest of the text to awkward to sorry edit any awkward paragraphs or sentences and otherwise search through the document for general issues. 01:39:14.680 --> 01:39:19.190 And we have a couple additional pages at the end for appendices of species characteristics. 01:39:23.310 --> 01:39:23.710 Great. 01:39:19.200 --> 01:39:25.470 If anybody has more interest in those as well, so that's a brief overview of our reports and PC. 01:39:25.480 --> 01:39:25.790 Thank you. 01:39:26.640 --> 01:39:27.760 Thanks so much, Jason. 01:39:29.220 --> 01:39:33.720 And if we can now pass it over to the Montana Natural Heritage program. 01:39:35.960 --> 01:39:36.510 Hey, Dan. 01:39:37.250 --> 01:39:37.790 How's it going? 01:39:38.970 --> 01:39:42.690 So yeah, this is great. 01:39:42.840 --> 01:39:47.200 Umm, you know, I think that I'm gonna talk about a lot of stuff. 01:39:47.210 --> 01:39:49.170 That was just talked about. 01:39:49.300 --> 01:39:53.540 Our reports are a little bit slimmer and I'll get into why we've kind of focused on that. 01:39:56.310 --> 01:39:56.880 Let's see. 01:39:56.890 --> 01:40:23.960 So what I really think about, you know what our reporting ohm products really focus on, it's really you know to get results out that can be to the land owners or land managers that occur within a cell, but also to the partners and volunteers, the folks who are putting out are detectors sure, not really have to tell anybody on this call, but acoustic data, particularly acoustic bat surveys can be really unsatisfying. 01:40:24.140 --> 01:40:25.990 You don't get the whole critters in hand. 01:40:26.000 --> 01:40:27.560 You kind of just put out electronics. 01:40:29.260 --> 01:40:29.990 Leaving there. 01:40:30.040 --> 01:40:30.550 Come back. 01:40:30.560 --> 01:40:31.230 Pick him up. 01:40:31.500 --> 01:40:43.170 You might get a percent of your card filled, so we want to really connect the people who are doing these surveys with with what the actual species are, that they're likely encountering. 01:40:43.900 --> 01:40:55.820 We also do things like produce interim reports, which I won't really talk about in depth here, as well as push our data to some of our web tools such as our map viewer and Environmental Summary report. 01:40:55.830 --> 01:41:25.600 And that's really the primary products we use to get those data in the hands of the land managers and folks who are who are thinking about where species are on the landscape, umm, and our goals for particularly for the products to talk about today are really to provide information on what species are presence within a cell and to kind of do some cheerleading, increase knowledge and excitement about the bats that are found within these cells and through this to encourage both future participation of land owners, we want to come back every year. 01:41:25.660 --> 01:41:30.090 So how do we get them excited about that as well as folks putting out detectors? 01:41:32.700 --> 01:41:36.660 So we kind of break down into two different products. 01:41:36.670 --> 01:41:38.390 To do this, we do a letter. 01:41:39.320 --> 01:41:52.360 This is a formal letter targeted at landowners, just basically detailing what we did, why we did it, and then where they can find further information on bats, again, to kind of get them excited. 01:41:52.700 --> 01:41:56.550 We also produce templated reports for those were curious. 01:41:56.560 --> 01:42:00.750 This is generated out of our SQL database using a Python script. 01:42:00.840 --> 01:42:03.160 Brayden can probably answer some questions on that. 01:42:03.170 --> 01:42:20.620 If you have any, but we've taken a little bit different approach, we try to keep these really slim, understandable and just due to actually primarily budget cycles where we're not starting or finishing up the work till after we've started next summer surveys, we want to get these out sooner. 01:42:20.630 --> 01:42:27.640 So we've just chosen to report auto classification data with a big kind of caveat that you might not be seeing. 01:42:27.870 --> 01:42:45.740 All the species reported might not actually be present, and we do cut off lower number of call sequences, but we kind of rely on that initial X blurry explanation and given that this isn't, these reports aren't designed to have anybody making decisions off them, there's better products for that. 01:42:45.750 --> 01:42:51.220 We feel pretty comfortable with that to give you a example of each. 01:42:52.030 --> 01:42:54.140 Again, we keep our reports really simple. 01:42:54.150 --> 01:42:57.990 You can see over here we have just some general information about the. 01:42:59.430 --> 01:43:01.500 Number of call sequences that were reported. 01:43:01.830 --> 01:43:14.100 We have our auto class calls again like Kathy was saying, we don't really do the scientific name for the species, but you can see how in this case it's a bunch of species. 01:43:14.110 --> 01:43:17.700 There's probably some in here that actually didn't occur, just given the potential confusion. 01:43:17.790 --> 01:43:19.910 Particularly are myotis species. 01:43:20.440 --> 01:43:22.790 We attach this to the our formal letter. 01:43:22.800 --> 01:43:28.670 We try to personalize these and address them to the folks that gave us permission. 01:43:28.740 --> 01:43:44.090 This can actually create some confusion because we're primarily getting address data from our tax data, and you can imagine if families lived in an area of eastern Montana for five generations, you have all sorts of different variations of who actually owns the land. 01:43:44.510 --> 01:43:48.050 Umm, you know who is who's managing the land? 01:43:48.060 --> 01:44:03.070 The person got permission from might not actually be the owner who might live in another state, so we're trying to sort this out in a database format, but we're still kind of getting our legs under us even a few years in on getting these data archived and managed well. 01:44:04.010 --> 01:44:17.470 But we try to get them there, provide a link to our field guide which is really helpful for understanding what bats there were, and we don't necessarily just limit the number of species that they potentially observed here. 01:44:18.130 --> 01:44:20.530 We have this for all 15 species. 01:44:22.260 --> 01:44:38.040 Additionally Ohm, we are sending the technicians out this summer and provide these to our other volunteers, but the PDF for these various field guide or species from our field guide, you can actually print these off our website. 01:44:38.050 --> 01:44:40.520 There's a link down there at the bottom. 01:44:40.530 --> 01:44:42.280 We're also sending them out with posters. 01:44:42.340 --> 01:44:45.100 Again, this kind of comes back to this cheerleading aspect. 01:44:45.170 --> 01:44:52.100 We want to get people excited about that, so when we continue to ask year after year, can we come on your land and do survey work? 01:44:52.110 --> 01:44:54.300 They're like, yes, that's awesome. 01:44:54.580 --> 01:44:56.720 Not just like, why are you talking to me again? 01:44:57.070 --> 01:44:59.040 Which we have had some land owners. 01:44:59.050 --> 01:45:04.400 Do you know some folks are fine with like once or twice and it's just not something they're interested in doing. 01:45:04.410 --> 01:45:10.580 So we wanna kind of head that off and really, really get people interested and excited about what we're finding. 01:45:10.770 --> 01:45:16.700 So I guess with that, you know, I can take questions in the chat or a little bit later here. 01:45:18.580 --> 01:45:19.010 Great. 01:45:19.020 --> 01:45:20.000 Thanks, Dan. 01:45:20.070 --> 01:45:24.350 And I really like the the cheerleading approach too. 01:45:24.360 --> 01:45:27.910 I think that's that's important for us all to keep in mind. 01:45:28.440 --> 01:45:43.210 So we'll pass it to Ted for just a a brief chat on hand delivered posters and then we'll open it up after Bethenny provides some remarks to to broader questions. 01:45:44.410 --> 01:45:45.170 Can you hear me? 01:45:46.540 --> 01:45:47.300 I can hear you. 01:45:47.310 --> 01:45:48.020 I can't see you. 01:45:47.730 --> 01:45:48.780 Yeah, OK. 01:45:48.850 --> 01:45:52.040 Anyway, I I lost some functionality anyway, so it's OK. 01:45:52.130 --> 01:45:54.900 I just wanted to kind of follow on what Dan said. 01:45:54.910 --> 01:46:00.240 I don't not gonna be able share anything but like it looked like his bats of Montana poster for individual land owners. 01:46:00.310 --> 01:46:01.520 That's what I deliver to them. 01:46:01.530 --> 01:46:04.920 Pictures of the basis that they detected and then what years they were detected. 01:46:05.030 --> 01:46:16.700 And like when I go to them, I see those little it's just so you know, 8 1/2 by 11 sheet of paper laminated and they post it like at a organic farm or whatever and they they love it. 01:46:16.710 --> 01:46:18.970 And then when I show up there the next day like, this is great. 01:46:19.220 --> 01:46:25.960 We got questions about this during the year, so that cheerleading aspect is is really important and it's low, low effort for me to create those. 01:46:25.970 --> 01:46:28.120 So I just want to encourage people to do things like that. 01:46:28.130 --> 01:46:30.110 It really seems to pay a lot of dividends. 01:46:30.600 --> 01:46:35.020 We can move on and I'll have try to come back on, I guess, yeah, I'm going to leave and come back. 01:46:36.420 --> 01:46:36.880 Great. 01:46:36.890 --> 01:46:37.520 Thanks, Ted. 01:46:38.750 --> 01:46:46.260 And a big thanks to all of our panelists, both today as well as on Tuesday. 01:46:47.750 --> 01:47:05.360 For for sharing all of your workflows and your your tricks and approaches to making this whole process more efficient and transparent to the partners that that you all are working with and and your geographies. 01:47:05.630 --> 01:47:06.900 So thank you for that. 01:47:10.140 --> 01:47:12.250 So Bethany, I'll pass it along to you. 01:47:19.480 --> 01:47:20.560 Can't hear you yet? 01:47:33.280 --> 01:47:36.350 So Bethany, just an FYI, you're microphones muted. 01:47:41.000 --> 01:47:51.080 Oh, looks like she's having some struggles with teams too, so I'll I'll just jump on in until Bethany can join us. 01:47:53.570 --> 01:47:57.090 So what are some thoughts among the group? 01:47:58.120 --> 01:47:59.030 Umm ohh. 01:47:59.380 --> 01:48:32.730 I'd like I I I lost your screen but are are there any key thoughts or takeaways from from folks is we have just a few minutes to digest what we've just this speed dating process is as Shannon said to the complexities that we're all working with but how we can also leverage the knowledge and skills and work flows of others uh across the network. 01:48:45.910 --> 01:48:46.320 OK. 01:48:47.140 --> 01:48:48.050 Yeah, there she is. 01:48:46.390 --> 01:48:50.020 Sorry everybody, I could not get my teams to like. 01:48:50.060 --> 01:48:59.270 It was behind everything and I couldn't get to my mute button so to see the a little ideas and brainstorming here. 01:48:59.360 --> 01:49:01.190 So some thoughts that came up. 01:49:01.280 --> 01:49:14.950 So the idea being around this discussion is when we went through a lot of content, real rapid fire and there's a lot that probably like this really gave folks a picture of what's possible and what's out there. 01:49:14.960 --> 01:49:22.190 But perhaps the launchpad on where we can propel even further and so curious to hear one. 01:49:22.200 --> 01:49:22.780 Like what if? 01:49:22.790 --> 01:49:25.910 What are folks ideas on where we could go? 01:49:25.920 --> 01:49:28.020 Like what we want this process to look like. 01:49:28.990 --> 01:49:29.390 Umm. 01:49:29.440 --> 01:49:31.320 Perhaps even five years from now. 01:49:31.520 --> 01:49:39.450 But then what are like some tangible next steps that we can take in the near term that help get more of this into folks hands. 01:49:39.460 --> 01:49:48.160 So for example, if you do code, how do you want to receive the code and like how do you want that interaction to look like? 01:49:48.170 --> 01:49:50.210 How do you want to be able to access it? 01:49:50.630 --> 01:50:02.510 And if you don't, then what are some additional if you don't write in our or in Python or in any of the but umm, I work with any of the applications that were used today. 01:50:02.520 --> 01:50:15.130 What some education or information that we could share, perhaps and in what format or modality that would help people perhaps grow their own assets or skills or take advantage of what they can? 01:50:15.730 --> 01:50:20.890 One idea, two ideas that we've come up and then I want to turn it over to folks and hear what yours are. 01:50:22.160 --> 01:50:48.330 One is we are really seriously considering USGS trying to pull together these coded solutions and doing an actual format formal code review and publication that would have a citable DOI and that we could we could bundle in as vignettes in an in different software packages and that we could iterate that and iterate new releases through time. 01:50:49.140 --> 01:51:19.980 So that's like one thought on how we could a share this in a well documented, reviewed and transparent way and another that's even lower hanging fruit was to take the presentations or recordings from today and from Tuesday break it up into different sections with summary descriptions and put it up on the website that people can go back and revisit to dig into the sections that they had interest on and review those. 01:51:20.210 --> 01:51:25.040 And I and from those either slides or recording say like. 01:51:25.050 --> 01:51:33.520 OK, here's a great point of contact that I can reach out to to learn more and at least figure out like where am I starting point could be on this road. 01:51:35.060 --> 01:51:37.710 So those are just two examples. 01:51:38.000 --> 01:51:59.980 Some others could include in-depth trainings like I was kind of thinking of something like Umm Aveda thon, but like going through all of these steps right that that were presented today on what that might not the actual technique of manual vetting, but like, what are the processes and steps that we go through? 01:52:00.650 --> 01:52:08.240 So yeah, I'll leave it there and see what ideas which suggestion rain. 01:52:11.960 --> 01:52:16.760 The what you were talking about putting it on the website and like having more information for who to go to. 01:52:14.750 --> 01:52:17.530 Oh yeah, yeah. 01:52:19.550 --> 01:52:19.670 Cool. 01:52:20.850 --> 01:52:22.480 Uh, that's like the easiest one. 01:52:24.610 --> 01:52:28.850 Uh, it's so long as we can like post the everyone's cool with post in the content. 01:52:29.790 --> 01:52:31.760 Umm. Yeah. 01:52:31.770 --> 01:52:33.430 And thanks Sarah for clarifying that. 01:52:33.440 --> 01:52:57.650 If we do these code releases or software releases, that would have a full citation and anyone who's contributed to that would be on the author line, and so it's also a way to kind of perhaps get get more visible credit for these contributions that folks are making in solutions they're developing, and what other ideas or I guess, where do you tell us where you have more most interest? 01:52:58.430 --> 01:52:59.490 What do you want the most? 01:53:04.930 --> 01:53:05.130 Yeah. 01:53:07.630 --> 01:53:10.750 So, like hands on trainings for implementing some of this. 01:53:17.140 --> 01:53:18.730 Maybe Nymphon next year reign. 01:53:21.030 --> 01:53:31.350 So I'm not the Co chair of about working group anymore, but Jill is and so I can definitely put a bug in her ear for, you know, as we come up for an Impala. 01:53:31.950 --> 01:53:32.520 Mm-hmm. 01:53:49.020 --> 01:53:50.170 I guess I like the. 01:53:50.180 --> 01:53:56.370 I like the the idea of trying to share all the different codes so long as people are are willing to do that. 01:53:56.600 --> 01:54:01.210 And I I like the DOI, but maybe even as an interim solution. 01:54:01.220 --> 01:54:03.550 We they could just be posted out there, right. 01:54:05.710 --> 01:54:06.040 Umm. 01:54:03.560 --> 01:54:20.740 If people wanna put it on GitHub, sounds like Ben's doing a great job of documenting the code, which is really key, and I know what the echo clean we just went through a session over the last couple weeks of like the the biggest impediment, as I said, was getting people to get Python installed on their machine. 01:54:21.370 --> 01:54:22.000 Who is her? 01:54:22.010 --> 01:54:23.720 Not a coder and all this is it's. 01:54:23.600 --> 01:54:23.780 Yeah. 01:54:23.730 --> 01:54:24.840 It's tricky. 01:54:24.850 --> 01:54:27.500 We've had, like people who are like, have all kinds of skills doing this. 01:54:27.510 --> 01:54:36.200 So we went through and and provided recommendations right in the beginning of like how to run ECHO clean is like here how to get Python running on your machine. 01:54:36.290 --> 01:54:38.060 Once you get that, it's super easy. 01:54:43.190 --> 01:54:43.370 Yeah. 01:54:38.630 --> 01:54:44.620 So we we documented that part of it, but then like say, so we could build upon that saying like here's our Python section. 01:54:44.630 --> 01:54:47.900 First thing, do this second thing you want to. 01:54:49.090 --> 01:54:50.700 You know, use it for manual wedding. 01:54:50.710 --> 01:54:51.100 Great. 01:54:51.110 --> 01:54:52.580 You wanna use it now for uploading? 01:54:52.590 --> 01:54:55.270 That's the next like little package and trying to link those together. 01:54:55.280 --> 01:54:58.710 I mean it takes work, and we're gonna have to ask people to to do that. 01:54:58.720 --> 01:55:02.470 But I I think it would be an awesome resource if we could do that. 01:55:03.210 --> 01:55:03.410 Yeah. 01:55:10.600 --> 01:55:23.120 I could see like we need the documentation that anyone could follow, but I can also see how having people in a room altogether with their computers and like OK go here. 01:55:23.130 --> 01:55:23.500 Click here. 01:55:23.510 --> 01:55:28.710 We're going to download Python together and then you know like that, yeah. 01:55:38.560 --> 01:55:38.740 Yeah. 01:55:28.780 --> 01:55:42.850 I I it actually reminds me of of what Kathy was sharing and having the different forms of reports for land owners versus, you know, agency biologists. 01:55:43.060 --> 01:55:58.220 And so having having these different trainings or approaches or resources available for folks who aren't coders and and don't dive deeply into the computer science realm versus others who are. 01:55:59.090 --> 01:55:59.270 Yeah. 01:56:01.220 --> 01:56:22.160 I also had the thought of umm having someone like Ben come and do a detail at USGS and like work side by side with our developers to figure to pluck out which of these would be relevant to deploy like in the partner portal. 01:56:22.350 --> 01:56:27.100 So some of the QA, QC stuff I know that like, what could like? 01:56:27.150 --> 01:56:32.420 What would be relevant for that like state in place, location? 01:56:32.810 --> 01:56:43.740 And then how could we deploy it so that some of these things could just already be there and be a click of a button and umm, so I could see opportunity for all of these things. 01:56:43.750 --> 01:56:45.520 Yeah, go ahead, Kathy. 01:57:09.510 --> 01:57:09.710 Yeah. 01:56:46.930 --> 01:57:11.390 Yeah, I think that what you just said sort of help summarize my like what was swirling around in my brain like how do we lower the barriers to participation in a lot of the data side of things to folks that are not on this call, like people who don't aren't actively working their whole job isn't to like get NABat data processed and uploaded and reported on. 01:57:11.400 --> 01:57:15.540 But like just, you know, the people that feel like the whole thing is too complicated. 01:57:16.400 --> 01:57:16.580 Yeah. 01:57:15.550 --> 01:57:23.470 So I think if there's elements that can be automated in and brought back into the Nabat system, you know, I don't know. 01:57:24.130 --> 01:57:24.630 Moodily. 01:57:23.480 --> 01:57:26.530 I'm just trying to sort of rack my brain of what that would look like but. 01:57:27.310 --> 01:57:53.760 But it also makes me realize too that like the the Umm labor and cost like whether it's time or or acquiring assets to kind of build out and invest in the management tools or frameworks, the smaller scale you're operation is like the the lower the return on investment right but like for? 01:57:54.470 --> 01:58:03.920 But the a lot of the examples that we've seen, they're dealing with a lot of data and it's just creating all these efficiencies where that return on investment is really high. 01:58:04.790 --> 01:58:14.380 And so you're pointing Kathy, at how we, you know, whatever we can deploy in the partner portal is gonna benefit a large number of users. 01:58:14.390 --> 01:58:17.530 But it's a a centralized effort to build that out. 01:58:17.950 --> 01:58:23.120 But similarly we have the same thing with hubs and state level efforts. 01:58:23.130 --> 01:58:34.040 I mean, Montana is a really good example of what they're doing at a state level, and they're seeing a lot of benefit, but also it's for additional tax that it's not just for bats and it's working for them really well. 01:58:35.310 --> 01:58:41.180 And Ben is motivated to develop a lot of the solutions he has because of the number of partners he's serving. 01:58:41.190 --> 01:58:50.300 So I I just also wanna say like I think that there's it's not gonna make sense for individuals who are working on a small scale to build all of these things out. 01:58:50.310 --> 01:58:59.460 But if they're collaborating with other people locally to invest in something that provides efficiencies and benefits, that to me is a role for these. 01:58:59.470 --> 01:59:02.250 For hubs, uh and I could see it. 01:59:02.260 --> 01:59:03.600 A lot of benefit there so. 01:59:04.900 --> 01:59:08.750 Yeah, I guess one other thing I would just say that also came to mind. 01:59:08.760 --> 01:59:13.000 Like, are there things that people are doing differently that we would want standardized? 01:59:13.010 --> 01:59:14.080 Like, are there? 01:59:15.460 --> 01:59:18.480 You know, I've everyone's doing their QA QC a little bit different. 01:59:18.490 --> 01:59:20.830 Like I could see. 01:59:20.840 --> 01:59:26.080 NABat wanting to develop standards of QAQC or something like that where we can like. 01:59:26.740 --> 01:59:37.650 Streamline the I don't know the the sort of minimum requirements for for QA, QC or minimum requirements for. 01:59:41.910 --> 01:59:42.130 Yeah. 01:59:38.810 --> 01:59:45.600 I don't know what, but like you know that we are doing a lot of things differently and reinventing the wheel a lot in slightly different ways. 01:59:45.610 --> 01:59:50.740 And I feel like sometimes that doesn't matter, and sometimes it might be nice if things are being done in a similar way. 01:59:51.380 --> 02:00:06.460 So I think that to my mind for that, there was a lot of overlap on on not entirely, but the the overlap on those Venn diagrams were were pretty broad and that I don't think it would be too big of a lift. 02:00:06.470 --> 02:00:12.220 We discussed this on Tuesday, coming up with a checklist of here are things to look out for and at this stage in your process. 02:00:12.630 --> 02:00:20.190 But then we could build out coded vignettes and our or Python that we could point to from that checklist. 02:00:20.200 --> 02:00:24.290 That's like, OK, here is a package that's available that hits at these pieces. 02:00:25.210 --> 02:00:31.040 Ah, and so we could have a little I think there's opportunity to unify, but we don't actually have to. 02:00:31.160 --> 02:00:39.680 We're not gonna be able to generalize 100%, right, but we can provide clean code for this example. Umm. 02:00:39.510 --> 02:00:50.320 So, so so to keep consistent with with my role as time keeper and to respect the rest of everybody else's time today. 02:00:51.720 --> 02:01:13.380 I I think we have several next steps and action items moving forward, so we can standardize some elements of this of these efforts that are that are being done at a local scale, so that it's also transparent and available across the network, right? 02:01:13.910 --> 02:01:14.530 So. 02:01:14.600 --> 02:01:16.850 So that's something that I think we should. 02:01:16.860 --> 02:01:22.030 We should circle back around to and prioritize what those elements are. 02:01:22.770 --> 02:01:41.420 Also moving forward, you know we can work towards, we had a lot of support for getting some of this content on to the NABat monitoring.org website so that folks on this call and others can revisit it and and help to see some of those future discussions. 02:01:41.720 --> 02:01:51.960 But as a big thanks to to all the panelists and Ted Weller for helping to organize this today, and I thank you all so much.