1 00:00:10,200 --> 00:00:13,150 - [Announcer] Tools and Technology contributed talks room. 2 00:00:13,150 --> 00:00:16,930 We are about to have our last presentation 3 00:00:16,930 --> 00:00:19,360 from Melissa Czarnowski 4 00:00:21,210 --> 00:00:23,730 with the Connecticut Department of Energy 5 00:00:23,730 --> 00:00:25,280 and Environmental Protection. 6 00:00:25,280 --> 00:00:29,160 She'll be presenting on "Monitoring Stream Connectivity 7 00:00:29,160 --> 00:00:29,993 with Trail Cameras." 8 00:00:29,993 --> 00:00:32,750 - [Melissa] Hi everyone, my name is Melissa Czarnowski. 9 00:00:32,750 --> 00:00:35,200 I've worked for the Connecticut Department of Energy 10 00:00:35,200 --> 00:00:36,650 and Environmental Protection. 11 00:00:39,920 --> 00:00:42,150 And today I'm gonna be presenting on a project 12 00:00:42,150 --> 00:00:45,130 that we've been working on for the past five years, 13 00:00:45,130 --> 00:00:47,440 which is "Monitoring Stream Connectivity 14 00:00:47,440 --> 00:00:49,300 with Trail Cameras." 15 00:00:49,300 --> 00:00:52,100 I realized this is a forest ecosystem conference 16 00:00:52,100 --> 00:00:54,080 so I'm not entirely sure how much 17 00:00:54,080 --> 00:00:56,120 of this will apply to your work, 18 00:00:56,120 --> 00:00:58,050 but hopefully you will leave this presentation 19 00:00:58,050 --> 00:01:00,673 with some good ideas for your own projects. 20 00:01:02,670 --> 00:01:04,670 Before I get started I just wanna say 21 00:01:04,670 --> 00:01:07,250 that we recently published this article, 22 00:01:07,250 --> 00:01:09,980 A Novel Method to Evaluate Stream Connectivity 23 00:01:09,980 --> 00:01:13,660 Using Trail Cameras in River Research and Applications, 24 00:01:13,660 --> 00:01:15,510 which goes into depth about our methods 25 00:01:15,510 --> 00:01:18,540 while evaluating stream connectivity through pictures. 26 00:01:18,540 --> 00:01:22,320 If you wanna dig into our research after this presentation. 27 00:01:22,320 --> 00:01:24,420 And I just wanna acknowledge the main author 28 00:01:24,420 --> 00:01:26,280 Chris Bellucci and coauthors, 29 00:01:26,280 --> 00:01:28,470 Mary Becker and Corinne Fitting. 30 00:01:28,470 --> 00:01:31,220 And in addition, Allie Hibbard and Laura Robbins 31 00:01:31,220 --> 00:01:33,010 who have been instrumental in helping me 32 00:01:33,010 --> 00:01:34,660 with the field work. 33 00:01:34,660 --> 00:01:36,840 So I've provided the link to that article 34 00:01:36,840 --> 00:01:39,440 right here in this corner in case you're interested. 35 00:01:41,310 --> 00:01:43,690 Today I'm going to go over a brief background 36 00:01:43,690 --> 00:01:46,100 of why we decided to start using trail cameras 37 00:01:46,100 --> 00:01:48,380 to monitor stream connectivity, 38 00:01:48,380 --> 00:01:51,300 our method, how we evaluate stream flow 39 00:01:51,300 --> 00:01:52,960 from the pictures, 40 00:01:52,960 --> 00:01:56,070 the metrics we use to evaluate stream flow impairment 41 00:01:56,070 --> 00:01:58,533 and the few examples of what we've seen so far. 42 00:02:01,270 --> 00:02:04,240 So why monitor stream flow with pictures? 43 00:02:04,240 --> 00:02:08,040 Well, first and foremost, pictures speak a thousand words. 44 00:02:08,040 --> 00:02:12,020 All of these pictures were taken on August 21st, 2020. 45 00:02:12,020 --> 00:02:14,020 Two of these streams are completely dry 46 00:02:14,020 --> 00:02:15,830 as you can see on the left 47 00:02:15,830 --> 00:02:18,680 and two of these streams have flowing water on the right. 48 00:02:19,640 --> 00:02:21,430 The two streams that are completely dry 49 00:02:21,430 --> 00:02:24,440 are located within the vicinity of water withdrawals 50 00:02:24,440 --> 00:02:26,140 which leads me to my second point. 51 00:02:27,140 --> 00:02:29,260 We began this project to document the effects 52 00:02:29,260 --> 00:02:31,930 of water withdrawals on stream flow. 53 00:02:31,930 --> 00:02:35,240 One of deeps responsibilities is to regulate water usage 54 00:02:35,240 --> 00:02:38,490 over 50,000 gallons per day through a permanent program 55 00:02:38,490 --> 00:02:40,533 that was established in 1982. 56 00:02:41,530 --> 00:02:44,870 Any water diversion that existed prior to 1982 57 00:02:44,870 --> 00:02:46,500 was allowed to continue withdrawing 58 00:02:46,500 --> 00:02:48,890 as much water as they registered for 59 00:02:48,890 --> 00:02:51,000 without any kind of environmental assessment 60 00:02:51,000 --> 00:02:52,103 or public review. 61 00:02:52,950 --> 00:02:55,630 These are known as registered diversions. 62 00:02:55,630 --> 00:02:58,360 Today we're observing some of these streams dry up 63 00:02:58,360 --> 00:03:01,603 next to these diversions, like these two shown here. 64 00:03:02,700 --> 00:03:04,670 So we knew that they were a problem 65 00:03:04,670 --> 00:03:06,200 but we didn't really have a good handle 66 00:03:06,200 --> 00:03:07,730 on the scope of the problem. 67 00:03:07,730 --> 00:03:10,890 Like when does it go dry and for how long? 68 00:03:10,890 --> 00:03:13,480 And since stream gauges are few and far between 69 00:03:13,480 --> 00:03:14,980 we needed to figure out a better way 70 00:03:14,980 --> 00:03:16,803 to document stream flow differently. 71 00:03:17,840 --> 00:03:20,173 Our solution was to use trail cameras. 72 00:03:22,110 --> 00:03:24,910 One thing that I wanna emphasize is that photos are data 73 00:03:24,910 --> 00:03:27,180 that provide a record of conditions. 74 00:03:27,180 --> 00:03:29,050 The photos need to be clear, centered 75 00:03:29,050 --> 00:03:31,140 and without any obstructions. 76 00:03:31,140 --> 00:03:33,860 Without this we wouldn't be able to qualitatively 77 00:03:33,860 --> 00:03:37,263 and semi-quantitatively assess the stream flow. 78 00:03:39,950 --> 00:03:42,840 Stream connectivity is important for the ecological health 79 00:03:42,840 --> 00:03:45,760 of the stream and downstream waters. 80 00:03:45,760 --> 00:03:47,460 We define stream connectivity 81 00:03:47,460 --> 00:03:50,100 as hydrologically connected pools and riffles 82 00:03:50,100 --> 00:03:53,580 that link stream habitat along the longitudinal continuum, 83 00:03:53,580 --> 00:03:55,860 so that's upstream to downstream. 84 00:03:55,860 --> 00:03:58,620 While also recognizing the lateral dimension, 85 00:03:58,620 --> 00:04:00,590 the connection to the flood plane 86 00:04:00,590 --> 00:04:02,893 and vertical connection to groundwater. 87 00:04:05,240 --> 00:04:07,840 We focus on the rear end growth file period 88 00:04:07,840 --> 00:04:11,530 which is from July 1st through October 31st. 89 00:04:11,530 --> 00:04:14,720 These are the flows that are needed to sustain aquatic life. 90 00:04:14,720 --> 00:04:17,320 And also typically the season of highest conflict 91 00:04:17,320 --> 00:04:20,093 with human uses such as lawn irrigation. 92 00:04:21,970 --> 00:04:25,120 This diagram shows our method for how we go from pictures 93 00:04:25,120 --> 00:04:28,280 to metrics to help inform decision-makers. 94 00:04:28,280 --> 00:04:29,630 I'll go into detail on a few 95 00:04:29,630 --> 00:04:31,670 of these in the next couple of slides, 96 00:04:31,670 --> 00:04:34,440 but in general we select a location on a stream 97 00:04:34,440 --> 00:04:37,620 that captures at-least one referral pool sequence. 98 00:04:37,620 --> 00:04:39,180 When we deploy the trail camera, 99 00:04:39,180 --> 00:04:42,110 we set it's take one picture per hour. 100 00:04:42,110 --> 00:04:44,390 When we download the images back in the office 101 00:04:44,390 --> 00:04:47,260 we assign the stream connectivity category to each image, 102 00:04:47,260 --> 00:04:51,040 which I'll go into more detail in the next slide. 103 00:04:51,040 --> 00:04:52,760 And then based on those categories 104 00:04:52,760 --> 00:04:56,180 we calculate the average daily stream connectivity metrics 105 00:04:56,180 --> 00:04:59,360 which I will also go into more detail on. 106 00:04:59,360 --> 00:05:01,250 So this helps us to visualize the data 107 00:05:01,250 --> 00:05:03,653 and it helps us to inform stream management. 108 00:05:05,910 --> 00:05:08,050 We developed a six category system 109 00:05:08,050 --> 00:05:11,140 to describe the variations in stream connectivity observed 110 00:05:11,140 --> 00:05:13,540 using trail camera images. 111 00:05:13,540 --> 00:05:16,430 Category one is completely dry. 112 00:05:16,430 --> 00:05:20,193 Category two has some pools of standing water but no flow. 113 00:05:21,060 --> 00:05:23,120 Category three has a minimal flow 114 00:05:23,120 --> 00:05:25,560 in which some pools and riffles are disconnected 115 00:05:25,560 --> 00:05:28,223 and some habitat types are not accessible. 116 00:05:29,870 --> 00:05:32,590 Category four has flows with well-connected pools 117 00:05:32,590 --> 00:05:34,070 and riffles, 118 00:05:34,070 --> 00:05:37,320 Category five is where the flow fills the stream channel 119 00:05:37,320 --> 00:05:40,510 at or just below bank full discharge. 120 00:05:40,510 --> 00:05:42,510 And category six is when the flows 121 00:05:42,510 --> 00:05:45,593 are above the bank full discharge and into the flood plane. 122 00:05:46,760 --> 00:05:48,740 Once we download the pictures from the field 123 00:05:48,740 --> 00:05:51,493 we assigned each picture one of these categories. 124 00:05:54,110 --> 00:05:57,410 We then use the categorical data to calculate metrics 125 00:05:57,410 --> 00:06:00,240 that quantify stream connectivity. 126 00:06:00,240 --> 00:06:02,780 We developed 30 stream connectivity metrics 127 00:06:02,780 --> 00:06:04,173 to represent this data. 128 00:06:05,320 --> 00:06:08,160 The duration metrics represent a period of time 129 00:06:08,160 --> 00:06:10,540 and images associated with a category. 130 00:06:10,540 --> 00:06:11,490 For example, 131 00:06:11,490 --> 00:06:14,603 the average number of consecutive days in category one. 132 00:06:15,740 --> 00:06:19,210 The frequency metrics represent how often an images 133 00:06:19,210 --> 00:06:23,403 is in a category such as number of days in category one. 134 00:06:24,490 --> 00:06:27,430 The magnitude metrics provide a statistical summary 135 00:06:27,430 --> 00:06:31,660 of the category such as the average flow category, 136 00:06:31,660 --> 00:06:33,900 and the timing metrics describe 137 00:06:33,900 --> 00:06:36,520 when category occurs temporarily, 138 00:06:36,520 --> 00:06:39,220 such as the Julian day of the first observation 139 00:06:39,220 --> 00:06:40,453 in category one. 140 00:06:42,360 --> 00:06:46,040 So here's an example of the timing and duration metrics. 141 00:06:46,040 --> 00:06:48,040 Here, we have seven streams that we monitored 142 00:06:48,040 --> 00:06:49,500 with trail cameras. 143 00:06:49,500 --> 00:06:51,160 The bars represent the stream flow 144 00:06:51,160 --> 00:06:53,200 from July through October, 145 00:06:53,200 --> 00:06:54,680 the blue sections of the bars 146 00:06:54,680 --> 00:06:57,110 show when the stream was connected, 147 00:06:57,110 --> 00:07:00,220 the pink section show when the stream was disconnected 148 00:07:00,220 --> 00:07:03,760 and the red section show when the stream was completely dry. 149 00:07:03,760 --> 00:07:05,300 As you can see from this fig here, 150 00:07:05,300 --> 00:07:07,750 four of the streams that we monitored stayed connected 151 00:07:07,750 --> 00:07:09,550 throughout the study period, 152 00:07:09,550 --> 00:07:11,420 two of these streams jumped back and forth 153 00:07:11,420 --> 00:07:14,290 between connected, disconnected and dry. 154 00:07:14,290 --> 00:07:16,810 And one of these streams clearly had a gradual decrease 155 00:07:16,810 --> 00:07:18,780 in stream flow until they remained dry 156 00:07:18,780 --> 00:07:20,180 throughout the study period. 157 00:07:22,030 --> 00:07:25,460 Here's an example of one of the frequency metrics. 158 00:07:25,460 --> 00:07:27,960 The number of days from July through October 159 00:07:27,960 --> 00:07:30,380 in which the flow was dry. 160 00:07:30,380 --> 00:07:32,320 Here you can see that Chidsey Brook 161 00:07:32,320 --> 00:07:37,030 had the most number of days, almost 50, in which it was dry, 162 00:07:37,030 --> 00:07:39,560 followed by Mill River and Cobble Brook. 163 00:07:39,560 --> 00:07:43,820 The four other streams had no days in which the flow was dry 164 00:07:43,820 --> 00:07:46,773 as this was also represented on the previous figure. 165 00:07:48,620 --> 00:07:52,480 And here's an example of the magnitude metrics. 166 00:07:52,480 --> 00:07:55,860 The average flow from July through October. 167 00:07:55,860 --> 00:07:57,890 This figure shows that four of these streams 168 00:07:57,890 --> 00:08:00,880 had an average flow category of four, 169 00:08:00,880 --> 00:08:04,290 whereas the other streams had an average flow category 170 00:08:04,290 --> 00:08:05,913 between two and three. 171 00:08:08,490 --> 00:08:11,210 This metrics example is a little more complicated 172 00:08:11,210 --> 00:08:13,610 but I think it's worth explaining. 173 00:08:13,610 --> 00:08:16,050 Here we have the seven streams that we've been looking at 174 00:08:16,050 --> 00:08:18,310 in the previous examples. 175 00:08:18,310 --> 00:08:20,760 Each line represents the change in stream flow 176 00:08:20,760 --> 00:08:23,633 from July 1st through October 31st. 177 00:08:24,610 --> 00:08:28,570 The white areas of the bars represent connected flow 178 00:08:28,570 --> 00:08:31,973 and the gray areas of the bars represent disconnected flow. 179 00:08:33,100 --> 00:08:36,010 Now the blue sections of the lines, 180 00:08:36,010 --> 00:08:38,570 show when the USGS reference gauges 181 00:08:38,570 --> 00:08:40,253 had normal to high flows. 182 00:08:42,040 --> 00:08:43,830 The black sections of the lines 183 00:08:43,830 --> 00:08:47,313 show when the USGS reference gauges had low floats. 184 00:08:48,500 --> 00:08:51,460 So we would expect the blue sections of the lines 185 00:08:51,460 --> 00:08:53,820 to be in the white areas of the bars 186 00:08:53,820 --> 00:08:55,380 and the black sections of the lines 187 00:08:55,380 --> 00:08:57,363 to be in the gray areas of the bars. 188 00:08:58,830 --> 00:09:02,010 We mostly wanna focus on the blue sections of the lines 189 00:09:02,010 --> 00:09:04,460 being in the gray areas of the bars, 190 00:09:04,460 --> 00:09:07,460 since we are specifically looking for flow impaired streams. 191 00:09:09,340 --> 00:09:11,330 I just wanna point out a few examples 192 00:09:11,330 --> 00:09:12,480 of where this occurred. 193 00:09:14,870 --> 00:09:17,120 As you can see, these three streams 194 00:09:17,120 --> 00:09:19,810 were observed to have disconnected the stream flow 195 00:09:19,810 --> 00:09:21,720 when the USGS preference gauges 196 00:09:21,720 --> 00:09:24,000 had normal to high flows. 197 00:09:24,000 --> 00:09:25,540 Adding data to the conclusion 198 00:09:25,540 --> 00:09:27,763 that these three streams are flow impaired. 199 00:09:29,240 --> 00:09:31,650 And these are the same three streams that we've been seeing 200 00:09:31,650 --> 00:09:34,143 as flow impaired in the previous examples. 201 00:09:36,460 --> 00:09:38,700 Now, I just wanna take you through an example 202 00:09:38,700 --> 00:09:39,980 of one of our trail cameras 203 00:09:39,980 --> 00:09:43,490 located near a USGS stream gauge. 204 00:09:43,490 --> 00:09:46,010 This is a short period from September 7th 205 00:09:46,010 --> 00:09:49,360 through October 2nd of 2016, 206 00:09:49,360 --> 00:09:51,860 and if you pay attention to this little red star 207 00:09:51,860 --> 00:09:53,420 on the hydrograph, 208 00:09:53,420 --> 00:09:55,430 you'll see how it correlates with the stream flow 209 00:09:55,430 --> 00:09:56,920 in the picture. 210 00:09:56,920 --> 00:09:58,700 So here we have September 7th 211 00:09:58,700 --> 00:10:02,433 with a stream flow of 2.7 cubic feet per second. 212 00:10:05,890 --> 00:10:10,063 September 10th 2.20 cubic feet per second. 213 00:10:11,920 --> 00:10:13,787 And now it's dropping to 1.2 CFS, 214 00:10:14,710 --> 00:10:17,390 and you can see that clearly in the stream flow 215 00:10:17,390 --> 00:10:18,293 in the picture. 216 00:10:20,230 --> 00:10:22,943 It jumps up to 3.2 CFS, 217 00:10:23,940 --> 00:10:26,663 goes back down to 0.97 CFS, 218 00:10:27,860 --> 00:10:30,873 jumps up to 4 CFS after a rain event. 219 00:10:32,310 --> 00:10:35,133 Goes back down to 1.4 CFS. 220 00:10:37,050 --> 00:10:40,623 And ends with 2.4 cubic feet per second. 221 00:10:42,930 --> 00:10:45,530 As I mentioned at the beginning of this presentation 222 00:10:45,530 --> 00:10:48,160 before we began this project we didn't have a good handle 223 00:10:48,160 --> 00:10:50,803 on the scope of the issue of flow impaired streams. 224 00:10:52,090 --> 00:10:55,010 The department is required to identify flow impairments 225 00:10:55,010 --> 00:10:57,610 with the integrated water quality report, 226 00:10:57,610 --> 00:11:00,453 which is a report that goes to the EPA every two years. 227 00:11:01,360 --> 00:11:04,700 In 2014, we had identified 34 miles 228 00:11:04,700 --> 00:11:07,000 of flow impaired streams throughout the state. 229 00:11:09,230 --> 00:11:12,670 And in 2016, after we began this project 230 00:11:12,670 --> 00:11:16,423 we identified 159 miles of full impaired streams. 231 00:11:18,030 --> 00:11:21,770 And finally, as an added bonus to this trail camera project 232 00:11:21,770 --> 00:11:23,450 we get lots of wildlife images 233 00:11:23,450 --> 00:11:25,000 which is always fun to look at. 234 00:11:26,530 --> 00:11:28,910 All of the source code and data for this project 235 00:11:28,910 --> 00:11:33,170 are freely available and open source at this link right here 236 00:11:33,170 --> 00:11:38,170 github.com/marybecker/streamconnectivitymetrics. 237 00:11:40,550 --> 00:11:42,020 And with that I would just like to say, 238 00:11:42,020 --> 00:11:44,280 thank you for taking the time to listen 239 00:11:44,280 --> 00:11:48,653 to my presentation today and I will take any questions. 240 00:11:52,000 --> 00:11:54,460 - [Announcer] Do you see this as a monitoring technique 241 00:11:54,460 --> 00:11:56,830 you will be maintaining as a regular part of business 242 00:11:56,830 --> 00:11:57,893 moving forward? 243 00:12:01,230 --> 00:12:02,900 - [Melissa] Hi Jim. 244 00:12:02,900 --> 00:12:04,890 Yeah for the foreseeable future 245 00:12:04,890 --> 00:12:07,950 I think that we will be continuing monitoring 246 00:12:07,950 --> 00:12:09,273 with trail cameras. 247 00:12:10,430 --> 00:12:13,310 We are specifically looking at the impact 248 00:12:13,310 --> 00:12:16,670 of water withdrawals to near streams. 249 00:12:16,670 --> 00:12:20,563 So we're in the process of actually getting water use data. 250 00:12:21,420 --> 00:12:22,640 So it would be really interesting 251 00:12:22,640 --> 00:12:25,160 to see what the pictures look like 252 00:12:25,160 --> 00:12:27,293 when we have the water use data as well. 253 00:12:28,390 --> 00:12:30,280 So for the foreseeable future we will, 254 00:12:30,280 --> 00:12:33,343 but I'm not sure about a long-term project. 255 00:12:38,550 --> 00:12:40,240 - [Announcer] And John McCann asked, 256 00:12:40,240 --> 00:12:44,080 was there any consideration of putting staff plates 257 00:12:44,080 --> 00:12:47,773 in the field of view to track stage as well? 258 00:12:50,310 --> 00:12:54,100 - [Melissa] So we monitor probably about 40 streams 259 00:12:54,100 --> 00:12:55,950 throughout the state of Connecticut 260 00:12:55,950 --> 00:12:58,692 and we just don't have the resources 261 00:12:58,692 --> 00:13:00,610 to do that for all of the streams that we're looking at, 262 00:13:00,610 --> 00:13:04,103 but it is a good consideration. 263 00:13:09,780 --> 00:13:12,060 - [Announcer] And Sarah Nelson asked, 264 00:13:12,060 --> 00:13:13,590 do you think this would work well 265 00:13:13,590 --> 00:13:17,493 in high gradient streams/mountainous sites? 266 00:13:19,400 --> 00:13:21,950 - [Melissa] I think it would be kind of difficult 267 00:13:23,380 --> 00:13:26,350 just because a lot of high gradient sites 268 00:13:26,350 --> 00:13:29,040 have a lot of boulders and cobbles 269 00:13:29,040 --> 00:13:30,740 that might make it more difficult 270 00:13:30,740 --> 00:13:34,220 to see what the stream flow is actually like 271 00:13:34,220 --> 00:13:35,463 in the pictures, 272 00:13:36,510 --> 00:13:40,260 just from my experience of going out to the site 273 00:13:40,260 --> 00:13:44,560 and seeing it in person comparing it to the picture. 274 00:13:44,560 --> 00:13:46,883 If we have a really cobbly stream, 275 00:13:48,000 --> 00:13:50,620 we might rate the picture a little bit lower 276 00:13:50,620 --> 00:13:52,810 than what it actually is in person, 277 00:13:52,810 --> 00:13:56,363 just because that view is obstructed by all the cobbles. 278 00:13:58,340 --> 00:14:00,030 - [Announcer] Andy Wood is asking, 279 00:14:00,030 --> 00:14:03,043 do you have any issues with cameras being tampered with? 280 00:14:04,950 --> 00:14:07,870 - [Melissa] For the most part we've lucked out. 281 00:14:07,870 --> 00:14:11,830 We've had a couple cameras that have been stolen 282 00:14:11,830 --> 00:14:16,830 and a few cameras that have been attempted to be stolen. 283 00:14:17,050 --> 00:14:19,920 We lock all of our cameras with a lock 284 00:14:19,920 --> 00:14:22,590 and tether it to the tree. 285 00:14:22,590 --> 00:14:27,327 And we also put a little tag on it that says, 286 00:14:27,327 --> 00:14:31,320 "This camera belongs to the department for research purposes 287 00:14:31,320 --> 00:14:33,630 please don't tamper." 288 00:14:33,630 --> 00:14:35,340 I'm not sure if that works or not, 289 00:14:35,340 --> 00:14:38,363 but so far we've been pretty lucky. 290 00:14:43,400 --> 00:14:46,050 - [Announcer] And there's a comment from John McCann. 291 00:14:47,140 --> 00:14:49,660 It would be interesting to know if NDVI 292 00:14:49,660 --> 00:14:51,280 along the stream tracked 293 00:14:51,280 --> 00:14:56,023 with the connected/disconnected patterns you see. 294 00:14:58,020 --> 00:14:59,923 And it says maybe more of a comment. 295 00:15:01,650 --> 00:15:03,470 - [Melissa] Okay, thank you. 296 00:15:03,470 --> 00:15:05,610 - [Announcer] Jim Duncan has a question. 297 00:15:05,610 --> 00:15:06,820 This is really in the weeds, 298 00:15:06,820 --> 00:15:09,300 but how are you dealing with the storage 299 00:15:09,300 --> 00:15:10,940 and preservation of images 300 00:15:10,940 --> 00:15:13,923 after you have extracted the info you need from them? 301 00:15:15,020 --> 00:15:17,930 - [Melissa] That's a really good and important question 302 00:15:17,930 --> 00:15:22,080 and it's something that we are working on. 303 00:15:22,080 --> 00:15:26,930 So right now we luckily don't have a storage space issue 304 00:15:27,920 --> 00:15:28,963 on our network. 305 00:15:30,490 --> 00:15:33,850 So we've been able to save all the images 306 00:15:33,850 --> 00:15:38,490 but having a way to query all the data 307 00:15:38,490 --> 00:15:40,220 has been a little bit more challenging 308 00:15:40,220 --> 00:15:43,650 and that's something that we are working on. 309 00:15:43,650 --> 00:15:46,310 I know Chris Bellucci is on the call 310 00:15:46,310 --> 00:15:49,380 if he has any thing further to say to that. 311 00:15:52,860 --> 00:15:53,770 - [Chris] Yeah. Thanks, Melissa. 312 00:15:53,770 --> 00:15:55,950 I will just say that that is a really 313 00:15:55,950 --> 00:15:57,843 and important consideration, 314 00:16:00,440 --> 00:16:04,470 just like any continuous data where you collect hourly, 315 00:16:04,470 --> 00:16:06,170 for example stream temperature 316 00:16:06,170 --> 00:16:08,610 or anything else you guys might be involved with. 317 00:16:08,610 --> 00:16:11,523 It soon adds up very quickly and Melissa mentioned, 318 00:16:12,560 --> 00:16:15,160 at any given year we have 30 to 40 cameras 319 00:16:15,160 --> 00:16:18,890 collecting hourly images and it adds up 320 00:16:18,890 --> 00:16:23,890 and so I think if you're gonna pursue this type of thing 321 00:16:24,530 --> 00:16:29,530 you might think about having a server dedicated to it. 322 00:16:29,750 --> 00:16:31,410 Not only for storage of the pictures 323 00:16:31,410 --> 00:16:33,380 cause the pictures really are the data, 324 00:16:33,380 --> 00:16:36,200 but also for processing and querying 325 00:16:36,200 --> 00:16:40,410 and trying to do any analysis with the data, 326 00:16:40,410 --> 00:16:42,133 very important consideration.