1 00:00:03,816 --> 00:00:05,791 [LAUGH] All right, and 2 00:00:05,791 --> 00:00:11,620 that means we have to dismiss messages on both of our machines. 3 00:00:17,600 --> 00:00:22,385 And I'm really pleased to be introducing you to Shelby Weiss, 4 00:00:22,385 --> 00:00:25,320 who is a PhD student in geography. 5 00:00:25,320 --> 00:00:31,100 And she's gonna talk about modeling a particular topic area, which is wildfires, 6 00:00:31,100 --> 00:00:34,330 which is fascinating both for Oregon and where I'm from, Alaska. 7 00:00:34,330 --> 00:00:40,459 And that's the topic of where she is doing her research. 8 00:00:40,459 --> 00:00:44,650 Okay, so without any more delays, please come up and 9 00:00:44,650 --> 00:00:47,378 say more about your background and let. 10 00:00:47,378 --> 00:00:50,490 >> Yeah, great, I'm really pleased to be able to talk to you all today. 11 00:00:50,490 --> 00:00:55,779 I am a second year student PhD student, and I have been working on this 12 00:00:55,779 --> 00:01:01,380 project since I cane into the Earth, Environment, and Society program in 2018. 13 00:01:01,380 --> 00:01:05,060 This is really the first time I've presented some of this work, 14 00:01:05,060 --> 00:01:08,710 so it's exciting for me to be able to do that today. 15 00:01:08,710 --> 00:01:13,370 So I want to just start by kind of talking through my academic and 16 00:01:13,370 --> 00:01:19,050 professional background and how I kind of came into doing forest modeling. 17 00:01:19,050 --> 00:01:23,660 So I started out in wildlife biology. 18 00:01:23,660 --> 00:01:27,420 I did my bachelors at Colorado State University and 19 00:01:27,420 --> 00:01:31,120 I did a minor in applied statistics because I knew I wanted to do some 20 00:01:32,290 --> 00:01:35,040 quantitative ecology work moving forward in my career. 21 00:01:36,800 --> 00:01:40,580 And after I finished my bachelor's, 22 00:01:40,580 --> 00:01:44,030 I took an internship at senior national wildlife refuge. 23 00:01:44,030 --> 00:01:49,920 So I was really moving into wanting to do land management work and 24 00:01:49,920 --> 00:01:51,190 wildlife management work. 25 00:01:51,190 --> 00:01:54,400 And so I interned on an Upper Peninsula of Michigan, and 26 00:01:54,400 --> 00:01:55,640 that was a really cool opportunity. 27 00:01:56,690 --> 00:02:00,870 They do some interesting work up there with ecosystem restoration and 28 00:02:00,870 --> 00:02:02,400 forest restoration. 29 00:02:02,400 --> 00:02:04,660 And it's a really big landscape, it's 95,000 acres. 30 00:02:04,660 --> 00:02:05,180 And they have 31 00:02:06,860 --> 00:02:11,530 kind of this nice opportunity where they have lots of surveying that goes on and 32 00:02:11,530 --> 00:02:16,960 then the opportunity to kind of experiment with their management too. 33 00:02:16,960 --> 00:02:20,260 But I stayed there for about a year and a half and 34 00:02:20,260 --> 00:02:24,390 then I returned to grad school at Ohio State University. 35 00:02:24,390 --> 00:02:30,920 And so I focused for my master's on forest and wildlife management and thinking 36 00:02:30,920 --> 00:02:35,570 about ecosystem restoration and how we balance different priorities restoration. 37 00:02:35,570 --> 00:02:38,740 And this is where I really started thinking more about fire in 38 00:02:38,740 --> 00:02:40,840 ecological restoration, especially in forests. 39 00:02:42,850 --> 00:02:47,580 And so I really was interested in it as a tool as well as just 40 00:02:47,580 --> 00:02:50,251 a really fascinating process. 41 00:02:50,251 --> 00:02:55,428 And so after I finished my master's, I'm actually from St. Louis, 42 00:02:55,428 --> 00:03:01,060 Missouri originally, so after my master's I wanted to get closer to home. 43 00:03:01,060 --> 00:03:04,325 And so I took a position in the plant records department at 44 00:03:04,325 --> 00:03:06,236 the Missouri Botanical Garden. 45 00:03:06,236 --> 00:03:11,919 So it was a real departure in terms of topics that I had worked on before, 46 00:03:11,919 --> 00:03:16,105 botanical gardens to conservation. 47 00:03:16,105 --> 00:03:21,045 And this position was really working on databases and trying to assemble 48 00:03:21,045 --> 00:03:25,595 a lot of different data together to help come up with priorities for conservation. 49 00:03:25,595 --> 00:03:30,725 So it was really interesting and it was also a really nice taste of what 50 00:03:30,725 --> 00:03:35,862 kind of publicly available data is out there and how you can bring it together 51 00:03:35,862 --> 00:03:40,560 in creative ways and apply it to whatever question you might have. 52 00:03:40,560 --> 00:03:44,500 And so then after that I decided I really missed doing research. 53 00:03:44,500 --> 00:03:47,720 That's really where I want to take my career from now on, and 54 00:03:47,720 --> 00:03:50,039 so I started looking for PhD opportunities. 55 00:03:51,380 --> 00:03:56,109 And I came across this opportunity at PSU in the Earth, Environment, and 56 00:03:56,109 --> 00:03:57,339 Society program. 57 00:03:57,339 --> 00:04:02,139 And so this my position was advertised to work on this specific project, and 58 00:04:02,139 --> 00:04:07,014 the thing that really drew me to it, because I'm not a modeller by training, 59 00:04:07,014 --> 00:04:07,540 right? 60 00:04:07,540 --> 00:04:10,100 I don't come from a modeling background at all. 61 00:04:10,100 --> 00:04:14,551 And what really drew me to this work was just the questions that were 62 00:04:14,551 --> 00:04:19,242 driving this project and the ideas of, we can answer these really 63 00:04:19,242 --> 00:04:24,094 big questions about what's happening in these rural parts of Alaska and 64 00:04:24,094 --> 00:04:27,770 what may happen to them moving forward in the future. 65 00:04:27,770 --> 00:04:29,936 And so the questions really brought me and 66 00:04:29,936 --> 00:04:32,167 have been kind of keeping me interested ever since. 67 00:04:32,167 --> 00:04:38,570 So just a quick outline of what I'm gonna cover today. 68 00:04:38,570 --> 00:04:43,090 I'll do a very brief overview of just modeling in general, and 69 00:04:43,090 --> 00:04:46,920 then the model that I work with, which is LANDIS-II. 70 00:04:46,920 --> 00:04:51,890 I'm guessing there's probably a good amount of familiarity with modeling based 71 00:04:51,890 --> 00:04:56,080 on the audience, so I'll try not to bore you too much with background stuff. 72 00:04:56,080 --> 00:04:59,870 But please feel free to interrupt me if you want to clarify anything or, 73 00:04:59,870 --> 00:05:04,210 I don't know, back up and explain myself further. 74 00:05:04,210 --> 00:05:07,200 And then the second half I really want to talk about Alaska and 75 00:05:07,200 --> 00:05:12,040 how we're applying this forest model in Alaska. 76 00:05:12,040 --> 00:05:15,143 So the purpose models, zooming way out, 77 00:05:15,143 --> 00:05:19,679 a model is a representation of a system or a process, right? 78 00:05:19,679 --> 00:05:24,751 So they allow us to organize ideas, frame data, and 79 00:05:24,751 --> 00:05:31,800 allow us to explore hypothetical scenarios or real scenarios, right? 80 00:05:31,800 --> 00:05:38,479 And they allow us to make predictions and extrapolate across scale and time. 81 00:05:38,479 --> 00:05:43,540 And so I like to think especially about the landscape model that I work with. 82 00:05:43,540 --> 00:05:48,145 It's kind of this nice laboratory where we get to answer 83 00:05:48,145 --> 00:05:52,705 cool questions in ways that we wouldn't be able to necessarily on the ground. 84 00:05:52,705 --> 00:05:55,635 We can explore processes, right? 85 00:05:55,635 --> 00:06:00,915 Especially when we're doing work in ecology and thinking about climate change, 86 00:06:00,915 --> 00:06:04,595 we're really needing to think about no analog futures. 87 00:06:04,595 --> 00:06:08,051 So meaning that we can't expect the future to look like the past. 88 00:06:08,051 --> 00:06:12,435 And so modeling is a nice way to kind of get out to what are these potential 89 00:06:12,435 --> 00:06:15,840 futures that we might be seeing, and what are some scenarios and 90 00:06:15,840 --> 00:06:20,158 ways that we can manage within that potential future. 91 00:06:20,158 --> 00:06:22,870 And, yeah, and we can conduct experiments. 92 00:06:22,870 --> 00:06:27,283 This is something that was definitely cool to me, coming from a field ecology 93 00:06:27,283 --> 00:06:33,170 background where during my master's I did one project at one site, and I was able 94 00:06:33,170 --> 00:06:38,570 to say something about that site, but I couldn't necessarily extend that too far. 95 00:06:38,570 --> 00:06:40,130 It was more of a case study. 96 00:06:40,130 --> 00:06:45,050 It was on site, and we could say something about that site. 97 00:06:45,050 --> 00:06:50,330 Cool thing here is that we can run replicates, get landscapes and 98 00:06:50,330 --> 00:06:55,712 do lots of things that aren't practically possible in the field. 99 00:06:55,712 --> 00:06:58,981 And then we can compare across conditions and 100 00:06:58,981 --> 00:07:02,773 manipulate those conditions in interesting ways. 101 00:07:02,773 --> 00:07:06,870 So types of models, again I don't wanna talk down to you. 102 00:07:06,870 --> 00:07:09,593 It sounds like everyone's pretty familiar with models. 103 00:07:09,593 --> 00:07:13,348 But just the difference between analytical models and simulation models,. 104 00:07:13,348 --> 00:07:17,052 Analytical models having a closed form solution. 105 00:07:17,052 --> 00:07:19,718 They're able to be expressed as a function. 106 00:07:19,718 --> 00:07:23,395 As opposed to simulation models which are often complex. 107 00:07:23,395 --> 00:07:26,589 They use these mathematical relationships and 108 00:07:26,589 --> 00:07:32,350 logical operations to represent structures and behaviors, but they're dynamic. 109 00:07:32,350 --> 00:07:36,010 And so each time you're on them you may not get the same result, right? 110 00:07:37,105 --> 00:07:39,160 Actions and feedbacks. 111 00:07:39,160 --> 00:07:41,250 And so I work with, of course, simulation models. 112 00:07:41,250 --> 00:07:45,282 So that's mainly where I'm headed here. 113 00:07:45,282 --> 00:07:50,940 And so I wanted to provide some context about vegetation simulation models. 114 00:07:50,940 --> 00:07:51,900 There's a lot out there. 115 00:07:51,900 --> 00:07:54,280 There's lots of different flavors of them, 116 00:07:54,280 --> 00:07:56,620 and you can really get at different questions. 117 00:07:59,310 --> 00:08:03,150 And depending on the model you use You have different limitations, right? 118 00:08:03,150 --> 00:08:07,340 So I wanted to cover kind of the scope of vegetation simulation models. 119 00:08:07,340 --> 00:08:11,140 I mean, I'm sure I'm probably missing things too, right? 120 00:08:11,140 --> 00:08:14,910 There's a lot out there, but these are some common ones. 121 00:08:14,910 --> 00:08:18,056 The first dynamic global vegetation models, 122 00:08:18,056 --> 00:08:20,724 these tend to operate at large scales. 123 00:08:20,724 --> 00:08:24,172 They operate with climate models, and so 124 00:08:24,172 --> 00:08:29,001 you're able to capture feedbacks between vegetation and 125 00:08:29,001 --> 00:08:32,463 disturbance along with the atmosphere. 126 00:08:32,463 --> 00:08:36,116 And you're able to really look at the effects of climate change on vegetation 127 00:08:36,116 --> 00:08:37,530 and carbon and water cycles. 128 00:08:38,960 --> 00:08:41,470 And so you can think about questions like 129 00:08:41,470 --> 00:08:43,820 how does vegetation respond to climate change? 130 00:08:43,820 --> 00:08:45,760 Or how will the vegetation respond to climate change, 131 00:08:45,760 --> 00:08:50,620 and you can make some estimations about changes in carbon pools and 132 00:08:50,620 --> 00:08:55,010 flexes and so much more actually. 133 00:08:55,010 --> 00:08:57,230 There's also state and transition models. 134 00:08:57,230 --> 00:09:01,990 So these are pretty widely used as well. 135 00:09:01,990 --> 00:09:05,884 And these are pretty different because the user defines the states and 136 00:09:05,884 --> 00:09:07,470 the pathways between them. 137 00:09:07,470 --> 00:09:12,410 So it's really up to whoever's modeling to decide what are the pathways that 138 00:09:12,410 --> 00:09:16,589 you want to model and you're going to have this kind of pre determined possible 139 00:09:17,600 --> 00:09:20,000 states in between those paths. 140 00:09:20,000 --> 00:09:22,690 So they're usually not species level, 141 00:09:22,690 --> 00:09:26,740 you're going to be modeling kind of these different communities and 142 00:09:26,740 --> 00:09:32,442 different states and you have to define how the system is working. 143 00:09:32,442 --> 00:09:37,610 And it's nice because you can actually look at a wide range of vegetation types. 144 00:09:37,610 --> 00:09:39,500 So this is an example about sagebrush 145 00:09:41,060 --> 00:09:45,500 transitioning between juniper woodlands and grasslands. 146 00:09:45,500 --> 00:09:49,344 And you can also use them to test alternative hypotheses 147 00:09:49,344 --> 00:09:52,220 about how you think this results, okay. 148 00:09:52,220 --> 00:09:54,810 But like I said, it's predetermined. 149 00:09:57,740 --> 00:10:03,510 And then finally of course landscape models, this is a really large group. 150 00:10:03,510 --> 00:10:07,570 And this is where I'm operating with Landis. 151 00:10:07,570 --> 00:10:09,817 So these operate at large spatial and temporal scales. 152 00:10:09,817 --> 00:10:13,914 Just to kind of, I guess, provide an interest that could mean anything 153 00:10:13,914 --> 00:10:16,115 depending on who you're talking to. 154 00:10:16,115 --> 00:10:23,098 In Alaska our landscape is 12 million hectares, it's pretty large. 155 00:10:23,098 --> 00:10:28,881 And it's actually the largest I think that Landis has been applied to so far. 156 00:10:28,881 --> 00:10:31,419 I think like, when I was looking through other projects and things, 157 00:10:31,419 --> 00:10:34,440 it's like you can do over a hundred thousand hectares or something like that. 158 00:10:34,440 --> 00:10:37,546 And in Alaska we're really blowing that up. 159 00:10:37,546 --> 00:10:42,520 And you can use large temporal scales as well. 160 00:10:42,520 --> 00:10:46,380 These different models, of course differ from one another in terms of how they 161 00:10:46,380 --> 00:10:49,290 model processes and the level of detail that they simulate. 162 00:10:49,290 --> 00:10:54,238 They can be species level, which the model I work with is. 163 00:10:54,238 --> 00:10:58,919 And we can answer questions about the outcome of repeated 164 00:10:58,919 --> 00:11:01,660 stochastic spatial processes. 165 00:11:01,660 --> 00:11:06,790 So yeah, these models tend to be spatial in nature. 166 00:11:06,790 --> 00:11:08,890 And this is cool because it allows us, and 167 00:11:08,890 --> 00:11:11,711 you can look at these individual species level traits, 168 00:11:11,711 --> 00:11:15,146 you can really start thinking about those with no analog features. 169 00:11:15,146 --> 00:11:20,520 So, there's Landis on our diagram here. 170 00:11:20,520 --> 00:11:25,837 And so I will talk about LANDIS next, I'm just trying to give 171 00:11:25,837 --> 00:11:31,602 some background about how this model works and how you can use it. 172 00:11:31,602 --> 00:11:35,350 So LANDIS, the family of LANDIS models has been around for over 30 years. 173 00:11:36,600 --> 00:11:41,359 Widely used, as you can see from our map here, these are a map of the world 174 00:11:41,359 --> 00:11:46,278 LANDIS II, I believe, LANDIS II is being applied or has been applied. 175 00:11:46,278 --> 00:11:51,230 And LANDIS II, which is the version I work with, is over 20 years old. 176 00:11:51,230 --> 00:11:53,420 And the great thing about LANDIS II is that it's open source. 177 00:11:53,420 --> 00:11:58,920 So there's a whole user community that makes things really 178 00:11:58,920 --> 00:12:05,550 nice if you want to tweak things or you can download any extension. 179 00:12:05,550 --> 00:12:09,830 Anyone can, click on the links. 180 00:12:09,830 --> 00:12:12,730 And what LANDIS does is it simulates cohort succession. 181 00:12:12,730 --> 00:12:19,470 So it simulates how different cohorts of trees grow on the landscape and die and 182 00:12:19,470 --> 00:12:24,896 interact among different species and all sorts of processes. 183 00:12:24,896 --> 00:12:29,548 So, really, how succession happens is an emergent property of 184 00:12:29,548 --> 00:12:34,545 the species life history attributes, as well as the disturbance and 185 00:12:34,545 --> 00:12:39,300 how those species are dispersed throughout the landscape. 186 00:12:39,300 --> 00:12:42,120 There's not a single pathway that it will follow. 187 00:12:42,120 --> 00:12:47,330 And it responds dynamically to climate change if you have a climate change, 188 00:12:47,330 --> 00:12:52,080 as well as really any species that you put into it. 189 00:12:52,080 --> 00:12:55,440 And so this is again our landscape diagram right? 190 00:12:55,440 --> 00:13:00,180 And we can define within a single cell, a species age cohort. 191 00:13:00,180 --> 00:13:05,020 So in the cell I've given the example of, in the cell on the landscape we sent 192 00:13:05,020 --> 00:13:09,870 paper version of white spruce or here, with a range of ages. 193 00:13:09,870 --> 00:13:13,200 And they have, you know, a given biomass. 194 00:13:13,200 --> 00:13:18,060 So really we start with LANDIS defining what's in your domain, 195 00:13:18,060 --> 00:13:19,948 what's in each of these cells on the landscape. 196 00:13:19,948 --> 00:13:26,090 And you also need to define your life history 197 00:13:26,090 --> 00:13:31,410 attributes to decide what happened? 198 00:13:31,410 --> 00:13:32,957 So this is an example of an input file. 199 00:13:32,957 --> 00:13:37,496 Some examples of attributes, 200 00:13:37,496 --> 00:13:41,163 things like longevity, 201 00:13:41,163 --> 00:13:45,701 how long do your species live, 202 00:13:45,701 --> 00:13:49,715 when do they have seeds, so 203 00:13:49,715 --> 00:13:55,500 like whether something is [INAUDIBLE]. 204 00:13:55,500 --> 00:13:57,780 So you really have to do your homework and 205 00:13:57,780 --> 00:14:02,410 these are based on literature and, sometimes expert knowledge. 206 00:14:03,970 --> 00:14:09,780 And you're also going to define how species respond to disturbance. 207 00:14:09,780 --> 00:14:14,658 So we have prescribed regeneration, and do they sprout when they're disturbed, 208 00:14:14,658 --> 00:14:19,910 do they have a special type of cone called stratus cones 209 00:14:19,910 --> 00:14:22,400 that open when they're heated? 210 00:14:22,400 --> 00:14:26,550 So species that are stratinous, like black spruce and lodgepole pine, 211 00:14:26,550 --> 00:14:27,740 is another example. 212 00:14:27,740 --> 00:14:32,130 They have these cones that close until a fire comes through and then the code is 213 00:14:32,130 --> 00:14:37,570 open and they're able to disperse their seed even if the individual dies. 214 00:14:37,570 --> 00:14:42,320 So that's a neatattribute that we want to capture when we're running these models. 215 00:14:42,320 --> 00:14:45,682 And then also just fire tolerance, how do they tolerate shade, 216 00:14:45,682 --> 00:14:47,123 all those kinds of things. 217 00:14:47,123 --> 00:14:55,271 And so LANDIS also simulates disturbance which I've kind of been getting at. 218 00:14:55,271 --> 00:14:56,663 And the neat thing is, 219 00:14:56,663 --> 00:15:00,489 is you can run different disturbance types simultaneously. 220 00:15:00,489 --> 00:15:07,764 So you can have fire and wind and disease on the same landscape and interacting. 221 00:15:07,764 --> 00:15:11,437 There's also a lot of different extensions, 222 00:15:11,437 --> 00:15:16,822 I'll talk about them in a minute, but you could simulate harvest. 223 00:15:16,822 --> 00:15:22,010 So LANDIS is sometimes used by course managers to look 224 00:15:22,010 --> 00:15:26,070 at different harvest scenarios and for planning and that kind of thing. 225 00:15:26,070 --> 00:15:31,030 Fuels management, right if we're modeling fire and 226 00:15:31,030 --> 00:15:34,639 these disturbance events are stochastic and they depend upon probabilities. 227 00:15:35,820 --> 00:15:39,852 And like I said they can overlap. 228 00:15:39,852 --> 00:15:43,410 LANDIS is spatially explicit. 229 00:15:43,410 --> 00:15:47,890 So, things like fire can spread from cell, 230 00:15:47,890 --> 00:15:52,280 to cell seed dispersal is also spatially explicit. 231 00:15:52,280 --> 00:15:57,666 So, you have Something's growing here could disperse and seed to adjacent cells. 232 00:15:57,666 --> 00:16:02,074 So things like that that are important to capture the spatial dynamics of 233 00:16:02,074 --> 00:16:03,670 are spatially explicit. 234 00:16:03,670 --> 00:16:07,910 But then within a given cell things like tree and 235 00:16:07,910 --> 00:16:12,150 shrub establishment and growth, mortality, 236 00:16:12,150 --> 00:16:17,564 those spatially processes just occurred at the cell level. 237 00:16:17,564 --> 00:16:20,626 And LANDIS, like I was mentioning before, 238 00:16:20,626 --> 00:16:24,369 it has many different extensions that you can apply. 239 00:16:24,369 --> 00:16:26,606 So it really depends on the questions. 240 00:16:26,606 --> 00:16:30,232 The questions should drive what you're turning on in LANDIS and 241 00:16:30,232 --> 00:16:31,448 what you're using. 242 00:16:31,448 --> 00:16:34,821 And there's a lot of different options even within these, 243 00:16:34,821 --> 00:16:39,343 like we have multiple fire extensions, depending on what types of behavior and 244 00:16:39,343 --> 00:16:41,514 how complex you want to get with fire. 245 00:16:41,514 --> 00:16:44,466 Same with with any of these. 246 00:16:44,466 --> 00:16:51,873 So within the extensions you're only ever gonna have one succession extension, 247 00:16:51,873 --> 00:16:56,480 so that's really driving how the trees grow. 248 00:16:56,480 --> 00:17:03,397 And so there's multiple options there, depending on what you are prioritizing. 249 00:17:03,397 --> 00:17:06,521 There's the one that we use captures below ground dynamics because 250 00:17:06,521 --> 00:17:08,565 we're working in Alaska with permafrost. 251 00:17:08,565 --> 00:17:11,080 So getting that below ground stuff is really important. 252 00:17:11,080 --> 00:17:14,830 So that's the succession extension that we're using. 253 00:17:14,830 --> 00:17:19,577 However, you can have none or many disturbance extensions, just be tacking 254 00:17:19,577 --> 00:17:25,260 them on, right, and [LAUGH] then getting really overwhelmed for the results. 255 00:17:25,260 --> 00:17:27,962 And then there's also a whole suite of output extensions. 256 00:17:27,962 --> 00:17:32,314 And so these kind of control what the model spits out at you at each time step. 257 00:17:32,314 --> 00:17:36,890 And so you can say, I want it to tell me about, 258 00:17:36,890 --> 00:17:42,067 you could predefine, I was looking a little bit for 259 00:17:42,067 --> 00:17:45,933 a while at habitat quality for moose. 260 00:17:45,933 --> 00:17:50,926 And so there's a way that you can use a map extension 261 00:17:50,926 --> 00:17:55,703 to help you summarize your data and classify it. 262 00:17:55,703 --> 00:17:58,989 So that's what you use those for. 263 00:17:58,989 --> 00:18:03,250 And like I said, the great thing about LANDIS is that it's open source. 264 00:18:03,250 --> 00:18:06,170 You can download extension code, tweak it. 265 00:18:06,170 --> 00:18:11,520 Anyone can really create their own extension. 266 00:18:11,520 --> 00:18:13,460 Obviously, there's learning curves with it, right? 267 00:18:13,460 --> 00:18:17,270 And LANDIS they put on trainings every year, so they really tried to do some 268 00:18:17,270 --> 00:18:21,640 outreach and make sure that people are interested in learning about it. 269 00:18:21,640 --> 00:18:26,420 They can go to a training and learn how to use this model. 270 00:18:26,420 --> 00:18:29,030 So that's really the primer of LANDIS, and so 271 00:18:29,030 --> 00:18:31,877 now I'm gonna get into talking about Alaska specifically. 272 00:18:31,877 --> 00:18:35,550 So this is the name of the project that I'm working on. 273 00:18:35,550 --> 00:18:36,770 It's funded by NSF, 274 00:18:36,770 --> 00:18:41,810 and it's a large collaborative effort across six different institutions. 275 00:18:41,810 --> 00:18:44,362 So I'm really just a small piece of this, but 276 00:18:44,362 --> 00:18:47,061 I feel really excited that I get to be part of it. 277 00:18:47,061 --> 00:18:51,177 So our work is looking at shifts in species composition and 278 00:18:51,177 --> 00:18:55,129 carbon sourcing status due to fire and climate change. 279 00:18:55,129 --> 00:18:58,738 And so yeah, really broad, right, in interior Alaska, sure. 280 00:18:59,747 --> 00:19:02,323 >> You, >> Sure. 281 00:19:02,323 --> 00:19:08,466 >> Could you say something about how LANDIS is tested against data? 282 00:19:08,466 --> 00:19:13,274 In other words, you put in all the knowledge that you have building a model. 283 00:19:13,274 --> 00:19:21,570 Do you ever then subject the model to some test of how well you reproduce the model? 284 00:19:21,570 --> 00:19:22,930 >> Yeah, good point. 285 00:19:22,930 --> 00:19:27,260 I did definitely skip over the calibration steps, right? 286 00:19:27,260 --> 00:19:32,490 And so yeah, there is a process to calibrate the many parameters of LANDIS. 287 00:19:32,490 --> 00:19:37,540 So for example, species growth, 288 00:19:37,540 --> 00:19:42,880 we use field data for that, so we try to match the growth curves to the field data. 289 00:19:42,880 --> 00:19:46,082 So there's a whole calibration process across parameters. 290 00:19:46,082 --> 00:19:51,197 I think there was a curve I showed before that on one of the second slides I showed. 291 00:19:51,197 --> 00:19:55,538 I could try to find it as an example. 292 00:20:06,608 --> 00:20:07,563 Here. 293 00:20:09,358 --> 00:20:13,801 So yeah, these are examples of four different parameters that get at 294 00:20:13,801 --> 00:20:18,411 productivity response to temperature and water and things like that. 295 00:20:18,411 --> 00:20:23,619 And we use field data to do these are publicly available data sources. 296 00:20:23,619 --> 00:20:28,590 And then following that, yes, it's also a good idea to do some validation on 297 00:20:28,590 --> 00:20:33,561 the ground to make sure that the patterns you're seeing make sense, right, 298 00:20:33,561 --> 00:20:35,719 with [INAUDIBLE] correct, sure. 299 00:20:35,719 --> 00:20:39,654 >> Could be a big discussion topic, the differences between calibrating and 300 00:20:39,654 --> 00:20:43,152 how you have to be very careful not to kind of switch calibrating and 301 00:20:43,152 --> 00:20:46,165 testing and treat a model that's calibrated as tested. 302 00:20:46,165 --> 00:20:49,066 There's another major, [CROSSTALK] >> Yeah, absolutely. 303 00:20:49,066 --> 00:20:50,559 >> [LAUGH] >> Yeah I know, 304 00:20:50,559 --> 00:20:53,734 didn't mean to sound like I was putting those together. 305 00:20:53,734 --> 00:20:57,742 >> It's one thing to calibrate parameters, 306 00:20:57,742 --> 00:21:03,531 it's another thing to test the complicated model in total and 307 00:21:03,531 --> 00:21:06,658 see if it performs in total well. 308 00:21:06,658 --> 00:21:09,263 Because you have all these emerging properties, but- 309 00:21:09,263 --> 00:21:12,067 >> [LAUGH] Yeah, and hopefully, 310 00:21:12,067 --> 00:21:15,206 as I get into the Alaska example, 311 00:21:15,206 --> 00:21:20,490 I can highlight some things there to talk more about it. 312 00:21:23,021 --> 00:21:24,129 Let m find my way back. 313 00:21:26,744 --> 00:21:28,723 Went too far. 314 00:21:33,120 --> 00:21:39,735 Cool, so here is Alaska, [LAUGH] right? 315 00:21:39,735 --> 00:21:44,334 So I just took like ten to give a little bit of background as to what's 316 00:21:44,334 --> 00:21:46,202 motivating this project. 317 00:21:46,202 --> 00:21:48,004 We're interested in boreal forests. 318 00:21:48,004 --> 00:21:51,542 Boreal forests are the world's largest terrestrial biome. 319 00:21:51,542 --> 00:21:58,181 30 to 50% of global carbon stocks and that's actually probably an underestimate. 320 00:21:58,181 --> 00:21:59,905 >> [INAUDIBLE] >> Yeah, 321 00:21:59,905 --> 00:22:03,125 we're increasingly discovering that. 322 00:22:03,125 --> 00:22:08,235 Yeah, we're probably under estimating how much carbon is stored in the north. 323 00:22:08,235 --> 00:22:12,231 And I think the big reason why we're probably under estimating it is because 324 00:22:12,231 --> 00:22:13,674 a lot of it is below ground. 325 00:22:13,674 --> 00:22:15,504 So it's not in the canopy. 326 00:22:15,504 --> 00:22:20,304 It's in the soil and tied up in frozen soils, and 327 00:22:20,304 --> 00:22:25,232 so it's understandably difficult to estimate. 328 00:22:25,232 --> 00:22:30,507 So this is the landscape that we're modeling with LANDIS, 329 00:22:30,507 --> 00:22:33,998 and the blue line is the Arctic Circle. 330 00:22:33,998 --> 00:22:39,380 And as I mentioned before, we're talking about 12 million years. 331 00:22:39,380 --> 00:22:45,730 And so within the boreal forest, 30 to 40% is black spruce. 332 00:22:45,730 --> 00:22:49,617 Black spruce is a pretty big species in Alaska. 333 00:22:49,617 --> 00:22:53,250 It has those serotonious cones that I mentioned before. 334 00:22:53,250 --> 00:22:57,506 And this is not encompassing of all the Alaskan forest types. 335 00:22:57,506 --> 00:23:00,297 I'm just trying to give a flavor here. 336 00:23:00,297 --> 00:23:03,510 But the other sort of unique thing about black spruce forests, 337 00:23:03,510 --> 00:23:07,754 whether they occur on north-facing slopes or within more of a wetland situation, 338 00:23:07,754 --> 00:23:10,135 is they're often underlined by permafrost. 339 00:23:10,135 --> 00:23:15,537 And so within this region, we're actually in the discontinuous permafrost sense. 340 00:23:15,537 --> 00:23:18,699 So you're not gonna see permafrost below ground everywhere on 341 00:23:18,699 --> 00:23:22,180 the landscape like you would in the tundra farther north. 342 00:23:22,180 --> 00:23:27,932 It's really going to be kind of interrupted and driven a lot by topography 343 00:23:27,932 --> 00:23:33,151 and vegetation and ground cover and a whole bunch of other things. 344 00:23:33,151 --> 00:23:37,738 And so on these kind of cooler wetter landscape positions, 345 00:23:37,738 --> 00:23:43,335 you often find black spruce, as opposed to our more deciduous species, 346 00:23:43,335 --> 00:23:47,576 like virgin Aspen as well as white spruce too. 347 00:23:47,576 --> 00:23:51,820 And black spruce has really been dominant on the landscape. 348 00:23:51,820 --> 00:23:56,573 For the past 6,000 years and fires play a key role in maintaining black 349 00:23:56,573 --> 00:24:01,566 spruce dominance, it's really well adapted to regenerate following fire. 350 00:24:01,566 --> 00:24:04,706 It has several competitive advantages like serotiny. 351 00:24:04,706 --> 00:24:07,527 It's able to grow on permafrost so 352 00:24:07,527 --> 00:24:12,013 obviously doesn't need a huge amount of soil to grow. 353 00:24:12,013 --> 00:24:16,923 And it can germinate on thicker organic soils which other species in 354 00:24:16,923 --> 00:24:18,998 this region don't prefer. 355 00:24:18,998 --> 00:24:21,686 So it has these different competitive advantages that have really 356 00:24:21,686 --> 00:24:23,080 given it the leg up historically. 357 00:24:24,170 --> 00:24:27,530 However, historically fires took place every 50 to 150 years. 358 00:24:28,655 --> 00:24:31,790 So those periods in between were really important to allow 359 00:24:31,790 --> 00:24:35,960 black spruce to regenerate and grow into reproductive maturity. 360 00:24:35,960 --> 00:24:39,510 So this is just an example of a fire that was this past summer. 361 00:24:39,510 --> 00:24:41,573 There was a pretty crazy fire season up there. 362 00:24:41,573 --> 00:24:46,331 This picture was taken on the side of the road, 363 00:24:46,331 --> 00:24:52,924 looking out at the burned scar of this huge fire in the interior. 364 00:24:52,924 --> 00:24:53,847 >> It's so weird, too, 365 00:24:53,847 --> 00:24:56,578 because looking at these little puffs of smoke all over the place. 366 00:24:56,578 --> 00:24:59,732 It's not a big contiguous fire, it's all burning, but 367 00:24:59,732 --> 00:25:01,711 you only see this little evidence. 368 00:25:01,711 --> 00:25:03,890 And it's a really weird kind of forest fire [LAUGH]. 369 00:25:03,890 --> 00:25:05,130 >> Yeah, consistent smoking. 370 00:25:05,130 --> 00:25:09,603 And it think they talked about, when I was at their class, when they talked about how 371 00:25:09,603 --> 00:25:12,826 the fire was probably gonna continue burning underground. 372 00:25:12,826 --> 00:25:15,538 It's like a slow creeping ground fire in a lot of cases. 373 00:25:15,538 --> 00:25:20,295 Cuz there's a lot of organic matter in the soil that is smouldering, 374 00:25:20,295 --> 00:25:23,585 and still, so yeah it's an interesting fire. 375 00:25:23,585 --> 00:25:28,986 >> Triggered by lightning or by what? 376 00:25:28,986 --> 00:25:30,381 >> Yeah, lightning, yeah, I mean, 377 00:25:30,381 --> 00:25:31,891 you can have like human- >> Humans [LAUGH]. 378 00:25:31,891 --> 00:25:34,489 >> Human ignitions, right, [INAUDIBLE]. 379 00:25:34,489 --> 00:25:39,071 But because Alaska, where we're talking about is so remote, 380 00:25:39,071 --> 00:25:43,228 you're generally talking about lightning ignitions. 381 00:25:44,928 --> 00:25:48,724 And, as we all know, climate change is happening, right? 382 00:25:48,724 --> 00:25:52,420 And in Alaska, it's being felt particularly keenly. 383 00:25:52,420 --> 00:25:56,421 The northern latitudes are warming much more rapidly than other parts of 384 00:25:56,421 --> 00:25:57,085 the world. 385 00:25:57,085 --> 00:26:00,015 These are just from model projections, 386 00:26:00,015 --> 00:26:04,417 what they're anticipating in the coming century, right? 387 00:26:04,417 --> 00:26:08,639 Something on the order of a ten degree increase, celsius. 388 00:26:08,639 --> 00:26:13,108 So Alaska is standing to see a lot of change and 389 00:26:13,108 --> 00:26:16,880 it's already seeing a lot of change. 390 00:26:16,880 --> 00:26:20,824 And that has huge implications for fire, and how many ignitions we're seeing and 391 00:26:20,824 --> 00:26:22,780 how many fires we're seeing each year. 392 00:26:22,780 --> 00:26:27,920 So we're seeing this increase in fire frequency in Alaska. 393 00:26:27,920 --> 00:26:32,209 So this is a graph that shows not only how many fires per year, but 394 00:26:32,209 --> 00:26:33,963 how many of those are big. 395 00:26:33,963 --> 00:26:39,335 So the orange represents these large over 5,000 hectare fires. 396 00:26:39,335 --> 00:26:43,039 So we're getting more of them and more of them that are big. 397 00:26:43,039 --> 00:26:47,590 And so this follows that you're gonna see a lot of overlap. 398 00:26:48,690 --> 00:26:52,950 As you have more fire, the probability of those fires overlapping one another goes 399 00:26:54,230 --> 00:26:56,680 up, given that you have fuel there. 400 00:26:56,680 --> 00:27:01,359 So this map shows, I kind of wish it wasn't yellow just now, but 401 00:27:01,359 --> 00:27:05,182 the darker like orange and red are areas of overlap. 402 00:27:05,182 --> 00:27:09,722 And that's really what this project is focused on, is understanding what's 403 00:27:09,722 --> 00:27:13,734 happening in those areas of overlap that are sort of unprecedented. 404 00:27:13,734 --> 00:27:19,642 And that there's concern over seeing more of those going forward in the future. 405 00:27:19,642 --> 00:27:23,610 And then, also can't talk about Alaska without mentioning permafrost and 406 00:27:23,610 --> 00:27:24,914 all the dynamics there. 407 00:27:24,914 --> 00:27:27,557 So not only do climate change and fire interact, but 408 00:27:27,557 --> 00:27:30,162 they have huge implications for permafrost thaw. 409 00:27:30,162 --> 00:27:35,880 After a fire, everything's black, less reflectant, you have lower albedo. 410 00:27:37,160 --> 00:27:41,690 Which means a lot more heat being absorbed into the ground. 411 00:27:41,690 --> 00:27:45,678 And so that's a very direct short term effect. 412 00:27:45,678 --> 00:27:51,012 A lot of what keeps permafrost on the landscape has to do with how 413 00:27:51,012 --> 00:27:57,872 much organic matter is on top of it, what sort of species are growing on top of it. 414 00:27:57,872 --> 00:28:03,624 So all of these things in terms of aboveground vegetation, 415 00:28:03,624 --> 00:28:07,621 things like that, are interacting with fire and 416 00:28:07,621 --> 00:28:13,399 climate change to result in permafrost thaw following a fire. 417 00:28:13,399 --> 00:28:18,491 So our next few slides are gonna be dealing with this particular question 418 00:28:18,491 --> 00:28:23,166 which asks, how increasing fire frequencies alter successional 419 00:28:23,166 --> 00:28:27,356 trajectories in aboveground vegetation in the interior? 420 00:28:27,356 --> 00:28:29,019 This is from one of our field sites. 421 00:28:29,019 --> 00:28:34,433 We have a field crew that, or a field team that are based 422 00:28:34,433 --> 00:28:39,025 at UC Denver that are leading the actual work on 423 00:28:39,025 --> 00:28:44,356 the ground in Alaska, across a spectrum of reburns. 424 00:28:44,356 --> 00:28:48,734 So they're doing really cool work, so this is an image from a one-burn fire. 425 00:28:48,734 --> 00:28:52,763 As you can see, lots and lots of snags still standing, 426 00:28:52,763 --> 00:28:56,540 even maybe some live trees that survived the fire. 427 00:28:56,540 --> 00:28:59,199 Lots of material on the ground, 428 00:28:59,199 --> 00:29:04,430 you have some shoots that regenerated, or maybe sprouted up. 429 00:29:04,430 --> 00:29:06,399 This is following two fires, 430 00:29:06,399 --> 00:29:11,210 so, you no longer see those snags or any remnant live trees. 431 00:29:11,210 --> 00:29:16,084 Things are looking more open, we're seeing deciduous shrubs, 432 00:29:16,084 --> 00:29:19,740 or probably birch and saplings at this point. 433 00:29:19,740 --> 00:29:23,102 And just to note, each of these, I'm gonna show a third image here. 434 00:29:23,102 --> 00:29:26,558 All of these are 15 years after a fire event, so 435 00:29:26,558 --> 00:29:29,523 it's controlling for a time since fire. 436 00:29:29,523 --> 00:29:35,303 And so this is after three fires, having walked through this particular stand, 437 00:29:35,303 --> 00:29:39,564 you do see an occasional spruce sapling in the understory. 438 00:29:39,564 --> 00:29:42,040 But this is actually an aspen grove. 439 00:29:42,040 --> 00:29:47,630 So kinda interesting seeing what's happening on these free burn areas. 440 00:29:47,630 --> 00:29:49,513 >> How do you know how many burns there have been? 441 00:29:49,513 --> 00:29:55,817 >> So there's datasets through the Monitoring Trends in Burn Severity, 442 00:29:55,817 --> 00:30:00,514 it's a publicly available data [CROSSTALK] >> These are all burns 443 00:30:00,514 --> 00:30:01,996 within the last century then. 444 00:30:01,996 --> 00:30:07,421 >> Yes, they're all since, I think all of these are within the last 60 years. 445 00:30:07,421 --> 00:30:11,670 Yeah, yeah, thanks for asking. 446 00:30:11,670 --> 00:30:19,227 So yeah, we used prior perimeter datasets to help target where to do our sampling. 447 00:30:19,227 --> 00:30:23,101 And then this this graph comes from, of course, inventory dataset, 448 00:30:23,101 --> 00:30:28,380 looking at, Different sights, different plots across Alaska. 449 00:30:28,380 --> 00:30:33,743 Just showing that as you have more number of fires within the same period, 450 00:30:33,743 --> 00:30:37,046 number of stands of black spruce decreases. 451 00:30:37,046 --> 00:30:42,364 Free burns are, there's only three plots inside this dataset, right? 452 00:30:42,364 --> 00:30:44,949 But the trend is [INAUDIBLE]. 453 00:30:44,949 --> 00:30:48,475 And so part of the field work's efforts is to understand 454 00:30:48,475 --> 00:30:52,151 the mechanisms behind what we're seeing in terms of black 455 00:30:52,151 --> 00:30:55,463 spruce being less less present within these areas. 456 00:30:55,463 --> 00:30:57,220 So thinking about the mechanisms, 457 00:30:57,220 --> 00:31:00,214 I feel like I've probably touched on some of these already. 458 00:31:00,214 --> 00:31:04,476 But fire returning before black spruce is sexually mature. 459 00:31:04,476 --> 00:31:07,857 That organic layer becoming reduced, 460 00:31:07,857 --> 00:31:13,601 removing this competitive advantage that spruce might have had. 461 00:31:13,601 --> 00:31:18,281 As well as permafrost thaw, which allows for greater rooting depths, 462 00:31:18,281 --> 00:31:21,564 again, representing a competitive advantage. 463 00:31:21,564 --> 00:31:26,743 And so these are some mechanisms that are essentially playing out. 464 00:31:26,743 --> 00:31:31,690 And so we need to figure out how we can capture them in our model. 465 00:31:31,690 --> 00:31:33,329 So as I mentioned before, 466 00:31:33,329 --> 00:31:38,330 there's different succession extensions through LANDIS that you can apply. 467 00:31:38,330 --> 00:31:44,040 And none of them so far have permafrost dynamics captured with them. 468 00:31:44,040 --> 00:31:46,850 And that's a pretty, I hope I probably made that point clear, 469 00:31:46,850 --> 00:31:49,790 that that's Important dynamic to capture. 470 00:31:49,790 --> 00:31:52,270 There's a lot of interactions between above and below ground. 471 00:31:52,270 --> 00:31:56,880 So right now we are getting ready to go into the testing phase of this 472 00:31:56,880 --> 00:31:58,390 new extension for Alaska. 473 00:31:58,390 --> 00:32:03,682 So this succession extension couples together a soil model, 474 00:32:03,682 --> 00:32:08,463 which is DAMM and cNiP GIPL, which is a permafrost model 475 00:32:08,463 --> 00:32:12,449 that models temperature at multiple depths. 476 00:32:12,449 --> 00:32:16,710 And then SHAW, which is a hydrology model. 477 00:32:16,710 --> 00:32:20,884 So we're in the final phases of coupling these models together and 478 00:32:20,884 --> 00:32:22,459 it's really exciting. 479 00:32:22,459 --> 00:32:26,052 So we're about to go into more testing of it. 480 00:32:27,630 --> 00:32:30,994 And I don't think I mentioned it yet, but 481 00:32:30,994 --> 00:32:36,310 we're modeling that huge landscape out of four hectares. 482 00:32:36,310 --> 00:32:39,110 And so there are, as you might imagine, 483 00:32:39,110 --> 00:32:44,190 lots of inputs that you'd need to kind of compile and wrangle together. 484 00:32:44,190 --> 00:32:45,655 To get lambdas to run, 485 00:32:45,655 --> 00:32:49,250 you'd have to define what's on the ground when you start. 486 00:32:49,250 --> 00:32:53,287 Which is not a trivial task, especially when you're working in Alaska, 487 00:32:53,287 --> 00:32:55,510 where a lot of the data is fairly sparse. 488 00:32:56,660 --> 00:33:02,593 So this can be done in a number of ways, depending on what you have available. 489 00:33:02,593 --> 00:33:07,335 But we use forest inventory analysis data and compute that based 490 00:33:07,335 --> 00:33:11,235 on land cover maps to tell lambdas what's on the landscape at time zero. 491 00:33:12,325 --> 00:33:15,597 And then you also can put elevation maps. 492 00:33:15,597 --> 00:33:19,920 You will have to have, if you're modeling climate change, 493 00:33:19,920 --> 00:33:23,584 obviously, climate data, or even that, right? 494 00:33:23,584 --> 00:33:28,647 But we're using this particular data set from the scenarios and 495 00:33:28,647 --> 00:33:31,000 Arctic planning groups now. 496 00:33:31,000 --> 00:33:35,700 And they have some downscaled climate data, 20 kilometers. 497 00:33:35,700 --> 00:33:39,760 And we are pulling up these specific variables. 498 00:33:39,760 --> 00:33:44,580 Because we're running fire on the landscape, we have a few additional 499 00:33:44,580 --> 00:33:51,550 variables like wind speed and direction important to have fire be dynamic. 500 00:33:51,550 --> 00:33:54,410 And then you have a whole host of soil maps that are also in. 501 00:33:55,540 --> 00:34:01,320 And so this takes a lot of time to assemble these, as you might imagine. 502 00:34:01,320 --> 00:34:09,300 And yeah, with Alaska, really hunting down data sources can be a challenge. 503 00:34:09,300 --> 00:34:11,350 And so this is what your output can look like. 504 00:34:11,350 --> 00:34:18,270 You can have the biomass map for Black Spruce at time 30. 505 00:34:18,270 --> 00:34:22,970 You can also, when you're running fire, have fire maps come at you every year. 506 00:34:22,970 --> 00:34:30,332 So to have these different maps and you specify what times they come at you and 507 00:34:30,332 --> 00:34:35,510 then you can do different types of spatial analysis or 508 00:34:35,510 --> 00:34:39,234 you start answering your questions. 509 00:34:39,234 --> 00:34:43,370 And so I've wanted to show a few results from some 510 00:34:43,370 --> 00:34:46,800 initial runs we did with this pilot version of our extension. 511 00:34:46,800 --> 00:34:51,287 So I wanna be pretty careful and honest about what we're finding. 512 00:34:51,287 --> 00:34:54,670 This is very much preliminary from this pilot extension. 513 00:34:54,670 --> 00:34:57,150 It's not our final extension, by any means. 514 00:34:57,150 --> 00:35:00,330 But just to kinda capture where I'm hoping to go with this, 515 00:35:01,650 --> 00:35:03,450 I wanted to look at cells that had burned. 516 00:35:03,450 --> 00:35:06,692 And so what did they start out as? 517 00:35:06,692 --> 00:35:10,360 And then over time and 518 00:35:10,360 --> 00:35:15,430 depending on how many fires they experienced, what did they turn into? 519 00:35:15,430 --> 00:35:17,910 What was dominant at the end of the simulation? 520 00:35:17,910 --> 00:35:24,340 So if we had a cell that had mostly conifer biomass in it at the start, 521 00:35:26,080 --> 00:35:31,550 after one fire, over time, we see these are proportions. 522 00:35:31,550 --> 00:35:33,780 So at 25 years, 523 00:35:33,780 --> 00:35:40,070 something like 32% of those cells returned to being dominated by conifer. 524 00:35:40,070 --> 00:35:45,420 The light green is dominated by deciduous, the gray is now a non-forested condition. 525 00:35:45,420 --> 00:35:48,830 And so you can see it's sort of trending back to having 526 00:35:48,830 --> 00:35:52,070 a higher proportion of conifer dominance over time. 527 00:35:52,070 --> 00:35:55,820 And so this is one fire situation. 528 00:35:57,060 --> 00:36:00,710 On the other hand, if you begin with deciduous at the start, 529 00:36:00,710 --> 00:36:06,040 you're gonna end up with mostly deciduous in most of your cells. 530 00:36:06,040 --> 00:36:09,420 But there are actually somewhere conifers kind of intruding and 531 00:36:09,420 --> 00:36:14,070 able to gain a foothold, which is kinda interesting. 532 00:36:14,070 --> 00:36:21,760 However, by the time you did three fires, so we have conifer at the start. 533 00:36:21,760 --> 00:36:24,680 And it's really kind of flat lines in terms of 534 00:36:24,680 --> 00:36:28,580 how many cells returned back to what they were in the beginning. 535 00:36:29,910 --> 00:36:33,990 >> Doesn't that suggest that that is the natural state of things? 536 00:36:35,910 --> 00:36:38,900 >> So, after three fires. 537 00:36:38,900 --> 00:36:42,630 >> That going back millennia, on average, 538 00:36:42,630 --> 00:36:48,150 you've got a lot of burns, but so it's three fires within a timeframe. 539 00:36:48,150 --> 00:36:51,970 >> Yes, yes, yeah, it's within the 100 year simulation. 540 00:36:51,970 --> 00:36:54,880 So you're talking about a return interval that's 541 00:36:54,880 --> 00:36:58,320 relatively short compared to historical frequencies. 542 00:37:00,010 --> 00:37:03,720 >> With the three fires, how far apart are they? 543 00:37:03,720 --> 00:37:06,260 If they're every 25 years, they're- >> Yeah, 544 00:37:06,260 --> 00:37:09,448 I haven't pulled out fire return interval yet for these. 545 00:37:09,448 --> 00:37:13,820 But yeah, I mean, that's definitely important to control for as well. 546 00:37:13,820 --> 00:37:18,380 But yeah, we can assume that these three fires happened within 547 00:37:19,500 --> 00:37:23,282 probably less than 30 year return or full, if your thinking about it. 548 00:37:23,282 --> 00:37:28,100 But yeah, definitely, there's a lot, yeah. 549 00:37:28,100 --> 00:37:33,400 >> So for every 25 years those three columns could just be the same thing. 550 00:37:33,400 --> 00:37:37,080 But it looks like after a fire they're very similar numbers. 551 00:37:37,080 --> 00:37:40,261 >> Right. >> That lets me know when fires happen. 552 00:37:40,261 --> 00:37:41,855 It does say time since most recent fire. 553 00:37:41,855 --> 00:37:44,010 >> Yeah, so this is the last fire. 554 00:37:44,010 --> 00:37:46,585 >> Five years after the third fire. 555 00:37:46,585 --> 00:37:49,582 >> Yeah, this is the last fire, sorry, that's right there. 556 00:37:49,582 --> 00:37:53,125 >> [LAUGH] >> Sorry, yeah, yeah, thank you. 557 00:37:53,125 --> 00:37:56,982 Yeah, so this is the third fire, this is after the third fire. 558 00:37:56,982 --> 00:37:59,816 It should have three icons there or something. 559 00:37:59,816 --> 00:38:00,895 >> [LAUGH] >> Is there data? 560 00:38:00,895 --> 00:38:01,952 This is a wall- >> I was thinking the same thing. 561 00:38:01,952 --> 00:38:02,615 >> This all looks the same. 562 00:38:04,560 --> 00:38:07,710 >> Yes, so yeah, so I mentioned our collaboration with UC Denver and 563 00:38:07,710 --> 00:38:11,760 they have the plots that are on the ground. 564 00:38:11,760 --> 00:38:16,695 >> So it'd be nice if you could show- >> What they're finding too? 565 00:38:16,695 --> 00:38:18,714 >> Yes, right. >> Not even of the same things, 566 00:38:18,714 --> 00:38:23,680 just could you kinda present similar data based on historical records? 567 00:38:23,680 --> 00:38:29,208 >> Well, we can't go look and find a spot that's 75 years 568 00:38:29,208 --> 00:38:34,630 after three fires because we didn't have three fires. 569 00:38:34,630 --> 00:38:38,890 We don't have data for when we have three fires until 60 years ago, so 570 00:38:38,890 --> 00:38:40,958 I appreciate what you're doing. 571 00:38:40,958 --> 00:38:42,133 [LAUGH] >> Thank you, yeah, no, 572 00:38:42,133 --> 00:38:43,980 it's certainly a challenge, right? 573 00:38:43,980 --> 00:38:45,485 We have our field folks. 574 00:38:45,485 --> 00:38:50,178 And yeah, in future presentations I'll certainly wanna include what they're 575 00:38:50,178 --> 00:38:52,256 seeing too, in terms of patterns. 576 00:38:52,256 --> 00:38:54,450 But they're looking at their plots. 577 00:38:54,450 --> 00:38:59,044 They're looking at 15 years since fire and they're controlling for that. 578 00:38:59,044 --> 00:39:01,971 And they're able to look at seedling density. 579 00:39:01,971 --> 00:39:06,535 So when you think about seedlings, seedlings are your future forest. 580 00:39:06,535 --> 00:39:11,277 So it stands to reason that if you're finding fewer seedlings after 581 00:39:11,277 --> 00:39:13,938 a three burn of black spruce, I mean, 582 00:39:13,938 --> 00:39:19,175 you're not gonna have matured black spruce in 30 years in that spot. 583 00:39:19,175 --> 00:39:24,470 So there are different inferences that you can make based on the field data. 584 00:39:24,470 --> 00:39:29,310 And there's been studies looking at this, these processes and 585 00:39:29,310 --> 00:39:33,070 how they're changing in Alaska with reburns. 586 00:39:33,070 --> 00:39:36,089 And most of them capture at least a two burn situation. 587 00:39:36,089 --> 00:39:39,420 So I think our 588 00:39:39,420 --> 00:39:44,810 field folks are really the only one that I know of that are looking at three burns. 589 00:39:44,810 --> 00:39:47,880 They are a pretty small proportion of the landscape right now. 590 00:39:48,890 --> 00:39:53,718 So yeah, there's data on What's happening after two burns and 591 00:39:53,718 --> 00:39:59,133 there is documented shifts to deciduous dominance after even just two burns. 592 00:40:00,767 --> 00:40:05,284 So yeah, [LAUGH] anyway I'm trying to keep myself honest here, these aren't 593 00:40:05,284 --> 00:40:10,560 finalized results by any means, I'm very much in the weeds with this right now. 594 00:40:10,560 --> 00:40:15,111 But yeah, so the story with deciduous, when you start with 595 00:40:15,111 --> 00:40:20,132 deciduous after the third fire, no conifers coming back at all. 596 00:40:20,132 --> 00:40:23,012 But a little more open, a little more green, clear ground potentially. 597 00:40:23,012 --> 00:40:26,937 And this is what kind of makes me scratch my head and 598 00:40:26,937 --> 00:40:32,462 think I need to maybe go back and do some more calibration with growth. 599 00:40:32,462 --> 00:40:38,209 So these are biomass versus number of times burned, looking at black spruce and 600 00:40:38,209 --> 00:40:43,097 paper birch, and we see that paper birch is higher than black spruce 601 00:40:43,097 --> 00:40:47,160 across all of these different kind of configurations. 602 00:40:47,160 --> 00:40:52,130 This is 25 years post fire, this is 65 years post fire and 603 00:40:52,130 --> 00:40:55,580 so I think there's still some work. 604 00:40:55,580 --> 00:40:58,183 So this is kind of a caveat, right, trying to keep myself honest. 605 00:40:58,183 --> 00:41:02,281 I need to go back, especially once we get the new extension up and running and 606 00:41:02,281 --> 00:41:04,531 do additional calibration to make sure. 607 00:41:04,531 --> 00:41:07,682 It's very important that we get those growth curves right, 608 00:41:07,682 --> 00:41:11,280 that we're really capturing how these species are growing, yeah? 609 00:41:11,280 --> 00:41:16,300 >> So you mentioned earlier that Landis just has a number of 610 00:41:16,300 --> 00:41:20,140 their succession extensions. 611 00:41:20,140 --> 00:41:25,380 So that's keeping track of some number of species and 612 00:41:25,380 --> 00:41:28,602 information about them, and how they grow and that kind of thing? 613 00:41:28,602 --> 00:41:31,636 >> Yeah. 614 00:41:31,636 --> 00:41:35,173 >> So this is just kind of plug and play with Landis. 615 00:41:35,173 --> 00:41:40,829 So take one that people have used, not that you have to each time indicate 616 00:41:40,829 --> 00:41:45,621 individual the species and all their- >> Yeah, so 617 00:41:45,621 --> 00:41:49,216 you're saying, once someone's figured out the parameters for a species, 618 00:41:49,216 --> 00:41:51,844 you should be able to apply it across- >> I'm asking [CROSSTALK]. 619 00:41:51,844 --> 00:41:58,023 >> Okay, yeah, so not necessarily, especially if you're moving regions. 620 00:41:58,023 --> 00:42:01,610 I mean, the same species grows differently in different locations. 621 00:42:01,610 --> 00:42:06,570 And so we're like way north in Alaska, so we really need to use Alaska data 622 00:42:06,570 --> 00:42:09,810 to get those growth curves right, because things grow slower up there. 623 00:42:09,810 --> 00:42:11,281 There's a shorter growing season. 624 00:42:11,281 --> 00:42:14,250 >> Right, so they're not general, yeah. 625 00:42:16,010 --> 00:42:17,220 >> Yeah, by species. 626 00:42:17,220 --> 00:42:22,600 Yeah, you would want to be region specific to where you're basing your model. 627 00:42:23,880 --> 00:42:25,748 >> So I was learning about the word biomass. 628 00:42:25,748 --> 00:42:28,531 It seems like birch could just be bigger trees or 629 00:42:28,531 --> 00:42:33,110 have a different characteristic in terms of their biomass accumulation. 630 00:42:33,110 --> 00:42:37,590 >> Spruces are little wimpy, scrawny things, mostly black spruce, anyway. 631 00:42:37,590 --> 00:42:40,672 >> Yeah, yeah, it's a fair point. 632 00:42:40,672 --> 00:42:46,119 >> But that would be the reason to measure biomass and not number of trees. 633 00:42:46,119 --> 00:42:51,205 >> But in terms of growth, what does growth in biomass 634 00:42:51,205 --> 00:42:55,471 mean versus growth in- >> Yeah, I guess it just makes me wonder, 635 00:42:55,471 --> 00:42:58,120 just because they're very different kinds of trees. 636 00:42:58,120 --> 00:43:00,660 >> As opposed to like density or age. 637 00:43:01,680 --> 00:43:05,210 >> Well, the fact that the number is higher, does that mean that they grow 638 00:43:05,210 --> 00:43:09,790 faster, or does it just mean that they grow more biomass per year? 639 00:43:10,858 --> 00:43:14,482 [LAUGH] >> Yeah, yeah it's a fair point. 640 00:43:14,482 --> 00:43:19,074 I think the thing that needs discussion, I think I need to look 641 00:43:19,074 --> 00:43:24,107 a little bit more is just thinking about on the landscape as a whole, 642 00:43:24,107 --> 00:43:27,574 tending to see when black spruce is dominant. 643 00:43:27,574 --> 00:43:31,399 You have stands that are mostly black spruce and 644 00:43:31,399 --> 00:43:36,093 then sometimes you have stand that are mostly deciduous. 645 00:43:36,093 --> 00:43:39,630 You don't see a lot of intermingling. 646 00:43:39,630 --> 00:43:44,843 I think that's why I'm a little bit, don't want to make any claims, right. 647 00:43:44,843 --> 00:43:49,677 So obviously this work is ongoing.. 648 00:43:49,677 --> 00:43:52,031 We want to more completely represent species composition. 649 00:43:52,031 --> 00:43:54,709 We've really just captured five species so far and 650 00:43:54,709 --> 00:43:59,270 I'm working on incorporating a couple of others into our landscape. 651 00:43:59,270 --> 00:44:02,569 And we're gonna be using that fully coupled, 652 00:44:02,569 --> 00:44:06,047 it's called the DGS extension moving forward. 653 00:44:06,047 --> 00:44:09,502 And we're obviously wanting to compare trends 654 00:44:09,502 --> 00:44:13,050 under historic climate versus climate change. 655 00:44:13,050 --> 00:44:15,603 So what I've shown you today is just historic climate, too, 656 00:44:15,603 --> 00:44:16,494 so keep that in mind. 657 00:44:16,494 --> 00:44:21,138 And I was really just pulling out those burn cells to say, okay, 658 00:44:21,138 --> 00:44:23,670 we experienced three fires. 659 00:44:23,670 --> 00:44:24,890 What do you see? 660 00:44:24,890 --> 00:44:29,237 So what we're gonna be probably doing a little bit, we will be doing more, 661 00:44:29,237 --> 00:44:33,586 asking more questions, right, with climate change, things like that. 662 00:44:33,586 --> 00:44:38,508 And we're also going to be using a dynamic fire extension called SCRPPLE, 663 00:44:38,508 --> 00:44:41,330 which make responsive to climate change. 664 00:44:41,330 --> 00:44:44,237 So there's two different fire extensions, 665 00:44:44,237 --> 00:44:48,367 one is base fire that sort of, you just tell it probabilities and 666 00:44:48,367 --> 00:44:52,370 it stochastically places fire on the landscape. 667 00:44:52,370 --> 00:44:57,068 And so we want to make sure that we're actually able to model fire with 668 00:44:57,068 --> 00:45:01,590 SCRPPLE and model how that could change over time. 669 00:45:01,590 --> 00:45:04,960 And then we're also going to be looking into changes in 670 00:45:04,960 --> 00:45:07,085 carbon source/ sink status. 671 00:45:07,085 --> 00:45:12,420 So looking below ground as well, 672 00:45:12,420 --> 00:45:15,300 in terms of how much carbon is in the system? 673 00:45:15,300 --> 00:45:22,070 Will it continue to be a sink in the boreal or is there the potential for 674 00:45:22,070 --> 00:45:27,343 the [INAUDIBLE] section source under climate change. 675 00:45:27,343 --> 00:45:31,293 So I wanted to kind of end with some of my personal takeaways about simulation 676 00:45:31,293 --> 00:45:33,280 modeling with LANDIS so far, right? 677 00:45:33,280 --> 00:45:38,510 Like I said, I'm only partway through this, but just to sort of circle back 678 00:45:38,510 --> 00:45:43,740 to what lessons I've kind of taken away so far. 679 00:45:43,740 --> 00:45:45,020 Of course, know your question. 680 00:45:45,020 --> 00:45:48,022 I mean, this probably applies across any research situation, right. 681 00:45:48,022 --> 00:45:52,388 And are you using the right tool, talking about all those different choices and 682 00:45:52,388 --> 00:45:52,920 models? 683 00:45:52,920 --> 00:45:56,810 Is your tool answer the question that you want? 684 00:45:58,090 --> 00:45:59,653 Can and should you adjust your tool? 685 00:45:59,653 --> 00:46:04,471 In the case of Alaska, we have the ability to build this new extension which was 686 00:46:04,471 --> 00:46:08,349 great and allows us to more completely get at these processes. 687 00:46:09,630 --> 00:46:10,600 Know your system, 688 00:46:10,600 --> 00:46:15,070 just really important to make sure that we understand mechanisms that are going on. 689 00:46:15,070 --> 00:46:18,832 We have a lot of collaborators which makes this really nice. 690 00:46:18,832 --> 00:46:24,783 College of Science in Denver and University of Florida. 691 00:46:24,783 --> 00:46:26,220 I'll go through all my collaborators at the end. 692 00:46:26,220 --> 00:46:29,240 But we're able to work with them and really talk to them about what they're 693 00:46:29,240 --> 00:46:34,730 seeing in their field data and what they know from their expertise, 694 00:46:34,730 --> 00:46:38,430 and to help us with the modeling. 695 00:46:38,430 --> 00:46:39,720 Getting comfortable with messy data. 696 00:46:39,720 --> 00:46:42,830 I showed that huge map of input sources, 697 00:46:42,830 --> 00:46:46,110 which is just you're having to draw from what you can. 698 00:46:46,110 --> 00:46:48,100 Especially in Alaska where the data is, 699 00:46:48,100 --> 00:46:52,910 you kind of have to work with what you have. 700 00:46:52,910 --> 00:46:54,380 So it's been a nice lesson for me, 701 00:46:54,380 --> 00:46:59,390 in terms of kind of how to creatively meld together different things like that. 702 00:46:59,390 --> 00:47:02,560 Of course, understanding the limitations of your model. 703 00:47:02,560 --> 00:47:07,751 And then finally, understanding what is emergent versus what is just prescribed, 704 00:47:07,751 --> 00:47:11,628 something that's very important if you're going to be taking 705 00:47:11,628 --> 00:47:14,205 anything away from what you're seeing. 706 00:47:14,205 --> 00:47:18,231 So with that, this is the group that I'm lucky enough 707 00:47:18,231 --> 00:47:21,828 to work with across all of these institutions. 708 00:47:21,828 --> 00:47:29,185 And if we have time, yeah, there's some time for discussion or questions. 709 00:47:29,185 --> 00:47:32,185 >> There was one question someone asked, they said, 710 00:47:32,185 --> 00:47:36,290 did you say that you were using the motus fire dataset? 711 00:47:36,290 --> 00:47:39,635 >> No, I'm using the MTBS, 712 00:47:39,635 --> 00:47:44,358 monitoring transit burn severity. 713 00:47:44,358 --> 00:47:46,058 [INAUDIBLE] It's a product, yeah. 714 00:47:46,058 --> 00:47:46,937 >> Okay. 715 00:47:49,730 --> 00:47:52,768 >> I had a question if nobody else did, you're about to ask one. 716 00:47:52,768 --> 00:47:53,530 Go ahead, please. 717 00:47:53,530 --> 00:47:56,091 [LAUGH] >> You said you were modelling this sphere 718 00:47:56,091 --> 00:47:56,727 of the large area. 719 00:47:56,727 --> 00:47:59,541 Is there a reason you're modelling the whole area as one? 720 00:47:59,541 --> 00:48:02,346 Is there are a lot of system interactions? 721 00:48:02,346 --> 00:48:05,433 Cuz it might be easier- >> [LAUGH] 722 00:48:05,433 --> 00:48:07,201 >> Than small, right? 723 00:48:07,201 --> 00:48:08,825 I mean, obviously, Gary can apply the model shape. 724 00:48:08,825 --> 00:48:11,801 Is there a reason to do the whole area as one model? 725 00:48:11,801 --> 00:48:14,888 >> Yeah, so one reason is fire. 726 00:48:14,888 --> 00:48:20,191 We wanna make sure that we can allow for really big fires to happen, and looking at 727 00:48:20,191 --> 00:48:25,981 multiple fires interacting and overlapping throughout the extent of the simulation. 728 00:48:25,981 --> 00:48:30,518 That's the only reason that [INAUDIBLE] in my mind. 729 00:48:30,518 --> 00:48:34,766 We're gonna be kinda like simulating dynamic fire and out loud. 730 00:48:38,940 --> 00:48:43,340 >> Yeah, I don't know if it's like a question but 731 00:48:43,340 --> 00:48:46,751 I really interested in succession. 732 00:48:46,751 --> 00:48:47,717 >> Okay yeah. 733 00:48:47,717 --> 00:48:52,204 >> So, I mean I have an idea about being about [INAUDIBLE] 734 00:48:52,204 --> 00:48:55,165 development [INAUDIBLE] session. 735 00:48:55,165 --> 00:49:03,274 So I wanna know what is the succession layer requiring [INAUDIBLE]. 736 00:49:03,274 --> 00:49:06,632 >> Yeah, so it's expensive. 737 00:49:06,632 --> 00:49:11,683 The succession extension is going to be stimulating 738 00:49:11,683 --> 00:49:15,941 things like influence [INAUDIBLE] species. 739 00:49:15,941 --> 00:49:16,852 >> Okay. >> And so 740 00:49:16,852 --> 00:49:20,657 they can go about doing that a number of ways. 741 00:49:20,657 --> 00:49:25,362 There's another extension to Landis that I've never worked 742 00:49:25,362 --> 00:49:29,081 with that's more like photosynthesis driven. 743 00:49:29,081 --> 00:49:32,938 Ours, but it doesn't capture below ground dynamics. 744 00:49:32,938 --> 00:49:39,282 So the different extensions kind of captured different parts of that. 745 00:49:39,282 --> 00:49:43,180 >> But in general- >> Influenced growth of competitions- 746 00:49:43,180 --> 00:49:45,652 >> And then which is promote or 747 00:49:45,652 --> 00:49:47,917 inhibit other species. 748 00:49:47,917 --> 00:49:53,687 >> Yeah, so there's you have competition for just mortality over time as PCs, 749 00:49:53,687 --> 00:49:56,972 displace one another through time, just, 750 00:49:56,972 --> 00:50:02,689 a disturbance comes through it's going to capture based on the condition. 751 00:50:02,689 --> 00:50:05,467 What's most, what's going to regenerate? 752 00:50:05,467 --> 00:50:06,659 And how will it regenerate? 753 00:50:06,659 --> 00:50:11,910 So it's something, if we have a fire come through and there was blow over their, 754 00:50:11,910 --> 00:50:17,178 bullet you can scrub back because we know that Willow sprouts are older, right? 755 00:50:17,178 --> 00:50:21,090 >> So, then you know it's right after the fire, it's whatever. 756 00:50:21,090 --> 00:50:23,882 It sprouts right after the fire and then. 757 00:50:23,882 --> 00:50:24,664 >> Proceed in. 758 00:50:24,664 --> 00:50:25,338 >> Yeah. 759 00:50:25,338 --> 00:50:31,198 >> Yeah, so it's very much driven by like dispersal, 760 00:50:31,198 --> 00:50:37,734 so seed dispersal or survival or reproductive strategy. 761 00:50:37,734 --> 00:50:38,294 >> Okay, yeah. 762 00:50:38,294 --> 00:50:44,736 Learning about the ecology? 763 00:50:44,736 --> 00:50:46,140 >> Yeah. 764 00:50:46,140 --> 00:50:54,131 [LAUGH] >> What's the economic of fire? 765 00:50:54,131 --> 00:50:57,012 >> [LAUGH] >> Yeah. 766 00:50:57,012 --> 00:51:02,928 >> [LAUGH] >> [LAUGH] Yeah. 767 00:51:02,928 --> 00:51:08,192 >> I've heard, Sorry, go ahead I guess. 768 00:51:08,192 --> 00:51:12,960 Anything that causes a massive. 769 00:51:12,960 --> 00:51:15,124 >> Yeah, yeah. 770 00:51:15,124 --> 00:51:17,415 A new product to be. 771 00:51:17,415 --> 00:51:18,608 >> Accross different fields. 772 00:51:18,608 --> 00:51:19,432 >> Yes. 773 00:51:19,432 --> 00:51:24,291 >> Similar ways of thinking in technology. 774 00:51:24,291 --> 00:51:26,477 >> That's what we do here. 775 00:51:26,477 --> 00:51:29,410 >> [LAUGH] >> One of it. 776 00:51:29,410 --> 00:51:30,200 >> Yeah. 777 00:51:30,200 --> 00:51:35,135 >> I just was curious, cuz I've heard LANDIS described as an agent-based model. 778 00:51:35,135 --> 00:51:39,255 And it doesn't feel like it's being used in that fashion, even though it 779 00:51:39,255 --> 00:51:43,986 may be in some way you could say for sure, it has agent-oriented features I don't. 780 00:51:43,986 --> 00:51:47,286 Agent, it feels very cellular for sure. 781 00:51:47,286 --> 00:51:48,116 >> Mm-hm. 782 00:51:48,116 --> 00:51:51,197 >> And there's a relationship between cellular thinking and 783 00:51:51,197 --> 00:51:54,216 agent thinking, except that the cellular thinking may or 784 00:51:54,216 --> 00:51:56,513 may not have any agency associated with it. 785 00:51:56,513 --> 00:51:59,718 >> Right. >> But it seems that most landscapes have 786 00:51:59,718 --> 00:52:01,968 relatively little agency. 787 00:52:01,968 --> 00:52:04,798 There are a whole lot victims of their circumstances. 788 00:52:04,798 --> 00:52:05,650 >> Yeah, yeah. 789 00:52:05,650 --> 00:52:07,235 >> So there's the kind of a difference in my mind. 790 00:52:07,235 --> 00:52:08,340 >> Right. 791 00:52:08,340 --> 00:52:12,126 >> How much agency you attribute to your agents? 792 00:52:12,126 --> 00:52:16,866 >> [LAUGH] >> If they're responding probabilistically 793 00:52:16,866 --> 00:52:19,359 to their environment? 794 00:52:19,359 --> 00:52:20,965 My answer, I mean, yeah. 795 00:52:20,965 --> 00:52:24,083 >> [LAUGH] Maybe it's not a fair question but 796 00:52:24,083 --> 00:52:29,503 it just feels like it's a great framework for doing this kind of work. 797 00:52:29,503 --> 00:52:33,325 And I just wondered if I've heard that word and 798 00:52:33,325 --> 00:52:36,965 I wondered if that was just, I don't know. 799 00:52:36,965 --> 00:52:39,377 I maybe I heard it in error, I don't know. 800 00:52:39,377 --> 00:52:44,447 >> Yeah, I'm not sure, since I haven't really been thinking about it that way, 801 00:52:44,447 --> 00:52:46,792 but the one thing to keep in mind too, 802 00:52:46,792 --> 00:52:51,500 is when we're operating at like the grid cell it's species age cohorts. 803 00:52:51,500 --> 00:52:53,968 So we're not modeling individual trees. 804 00:52:53,968 --> 00:52:57,447 It's just, what's the population? 805 00:52:57,447 --> 00:52:59,337 How many- >> Yeah. 806 00:52:59,337 --> 00:53:02,074 >> Or more you're mildly biomass or you're imputing something. 807 00:53:02,074 --> 00:53:06,267 >> Yeah, it's biomass rather than, yeah. 808 00:53:06,267 --> 00:53:07,722 It's pretty interesting. 809 00:53:07,722 --> 00:53:12,573 My favorite advisor Melissa Lukash has done some work with virtual reality and 810 00:53:12,573 --> 00:53:17,203 has had to work with some virtual reality folks to figure out how to kind of do 811 00:53:17,203 --> 00:53:21,045 a crosswalk between biomass and, what would that look like? 812 00:53:21,045 --> 00:53:24,416 So it's sort of really neat application, yeah. 813 00:53:24,416 --> 00:53:29,151 >> Okay, so you're talking about, you're keeping track of an age cohort and 814 00:53:29,151 --> 00:53:30,042 the biomass. 815 00:53:30,042 --> 00:53:32,012 >> Mm-hm. 816 00:53:32,012 --> 00:53:39,061 >> Are you keeping track of at least sorta the age of the biomass? 817 00:53:39,061 --> 00:53:39,846 >> Yes. 818 00:53:39,846 --> 00:53:44,409 >> Okay, I mean, we know how long it's been since have burned. 819 00:53:44,409 --> 00:53:48,305 Well, then so it's not just like biomass over time is that some 820 00:53:48,305 --> 00:53:52,206 proportion of the biomass is in trees that are 20 years old and 821 00:53:52,206 --> 00:53:55,760 some proportion of it is in trees that are 19 years old. 822 00:53:55,760 --> 00:53:59,904 >> Yes, yeah, so it kind of it's- >> Does it matter? 823 00:53:59,904 --> 00:54:02,056 I guess is the question for. 824 00:54:02,056 --> 00:54:05,821 >> Yeah, yeah, I think it matters especially if you want to get it like 825 00:54:05,821 --> 00:54:09,873 temporal trends and specifically say that it's black spruce is great. 826 00:54:09,873 --> 00:54:13,922 Yeah, but like young black spruce is what the biomass [INAUDIBLE]. 827 00:54:13,922 --> 00:54:18,834 >> Well, and you mentioned I guess they have to be 30 years old to 828 00:54:18,834 --> 00:54:23,469 be sexually mature that matters a great deal how much of that 829 00:54:23,469 --> 00:54:27,751 stand is 30 years old versus not 30 years old, so. 830 00:54:27,751 --> 00:54:31,329 >> Right. >> So I guess yeah, how does that get, 831 00:54:31,329 --> 00:54:36,069 I'm asking because I need to figure that out also. 832 00:54:36,069 --> 00:54:38,345 >> [LAUGH] [INAUDIBLE] >> What do you think about this 833 00:54:38,345 --> 00:54:39,113 [INAUDIBLE]? 834 00:54:39,113 --> 00:54:41,218 >> [INAUDIBLE] >> You good? 835 00:54:41,218 --> 00:54:43,106 [LAUGH] >> I'm not sure. 836 00:54:43,106 --> 00:54:43,867 >> Okay. 837 00:54:43,867 --> 00:54:49,034 >> I'm trying to think about your question. 838 00:54:49,034 --> 00:54:54,052 >> [LAUGH] >> Did the south track the growths, or 839 00:54:54,052 --> 00:54:59,033 is it just like, this is the main problem 840 00:54:59,033 --> 00:55:04,323 with the south [INAUDIBLE] something else. 841 00:55:04,323 --> 00:55:10,075 Like how much data does each cell track towards 842 00:55:10,075 --> 00:55:16,279 the different biomass, like a this is what it is? 843 00:55:16,279 --> 00:55:18,678 >> Sort of how granular is it with- >> Yeah. 844 00:55:18,678 --> 00:55:22,851 So, yeah, when you set your starting condition, 845 00:55:22,851 --> 00:55:27,813 you're going to basically assign each grid cell a map code. 846 00:55:27,813 --> 00:55:31,764 And within that map code, you have sort of a line that says, 847 00:55:31,764 --> 00:55:35,575 you have 10-year-old black spruce at this biomass. 848 00:55:35,575 --> 00:55:40,128 20-year-old black spruce up this biomass, biomass. 849 00:55:40,128 --> 00:55:41,525 >> That's all in one. 850 00:55:41,525 --> 00:55:43,327 That's a map code. 851 00:55:43,327 --> 00:55:46,659 >> That's a map code that's associated with a cell on the land. 852 00:55:46,659 --> 00:55:49,263 >> Okay, so some granularity and within a cell, and 853 00:55:49,263 --> 00:55:52,623 maybe that's under your control to some degree, or maybe not. 854 00:55:52,623 --> 00:55:53,858 >> Yeah. 855 00:55:53,858 --> 00:55:55,169 >> [INAUDIBLE] the session. 856 00:55:55,169 --> 00:55:55,704 >> Mm-hm. 857 00:55:55,704 --> 00:55:56,706 >> In this session? 858 00:55:56,706 --> 00:56:00,439 >> Yeah, so how's fast those grow and 859 00:56:00,439 --> 00:56:04,677 how much biomass [INAUDIBLE] been there. 860 00:56:04,677 --> 00:56:09,072 >> So is that very early on your presentation [INAUDIBLE] stand and 861 00:56:09,072 --> 00:56:10,855 transition [INAUDIBLE]. 862 00:56:10,855 --> 00:56:12,168 >> Uh-huh. 863 00:56:12,168 --> 00:56:15,250 >> Is that what the succession model is doing, 864 00:56:15,250 --> 00:56:19,080 is saying something like from this state at this time, 865 00:56:19,080 --> 00:56:24,683 given this composition who sees it the next time it should be some composition? 866 00:56:24,683 --> 00:56:31,628 >> No, it's the same transition models aren't looking at those. 867 00:56:31,628 --> 00:56:38,149 Like they're not modeling like individual species dynamics like LANDIS says. 868 00:56:38,149 --> 00:56:38,842 >> Okay. 869 00:56:38,842 --> 00:56:41,532 >> So LANDIS could, might like those transition models 870 00:56:41,532 --> 00:56:44,775 can only really follow those pathways to those defined states. 871 00:56:44,775 --> 00:56:50,478 LANDIS could come up with a unique composition of the species. 872 00:56:50,478 --> 00:56:50,990 >> Okay. 873 00:56:50,990 --> 00:56:55,007 >> [INAUDIBLE] that succession is a [INAUDIBLE] on the way you're treating 874 00:56:55,007 --> 00:56:59,523 [INAUDIBLE] and the [INAUDIBLE] succession model is [INAUDIBLE], is that right? 875 00:56:59,523 --> 00:57:07,352 >> Yeah, I think that I would say that, yeah. 876 00:57:07,352 --> 00:57:09,868 >> Well, I'm looking at the clock, it's one, I think we should thank our speaker. 877 00:57:09,868 --> 00:57:11,716 >> [APPLAUSE] >> Okay. 878 00:57:11,716 --> 00:57:12,761 >> And you still got [INAUDIBLE] how to keep talking? 879 00:57:12,761 --> 00:57:15,240 Yeah, yeah, I should put my email up here, but 880 00:57:15,240 --> 00:57:19,093 I'll put my email up if people are interested in talking more about this. 881 00:57:19,093 --> 00:57:23,391 >> Okay, well, I'll go ahead and shut the recording off now. 882 00:57:28,049 --> 00:57:30,499 >> Consider what you mentioned that there's a potential for 883 00:57:30,499 --> 00:57:32,050 the to become a carbon source