Edward Plumb Okay, well, good morning, everyone, and welcome to this month's ACCAP webinar, Tracking Drought in Alaska, Developing New Indicators and Tools. So, today's event is being hosted by the Alaska Center for Climate Assessment and Preparedness. We're also known as ACCAP. And we're located here in the International Arctic Research Center on the University of Alaska Fairbanks Troth Yeddha Campus. ACCAP is part of the Regional Integrated Sciences and Assessments program, which is funded by the NOAA Climate Program Office. So, for nearly 20 years, ACCAP has worked with communities, tribal organizations, various agencies, researchers, and other partners across Alaska to provide weather and climate information that supports both preparedness, resilience, and adaptation to a changing environment. So, thank you all for joining this morning. So when we think about Alaska, drought may not be the first hazard that comes to mind. We tend to think Alaska as having plenty of water, but drought does occur here, and there have been significant impacts on communities, ecosystems, hydropower, subsistence resources, agriculture, wildfire, and water supply. One of the challenges is figuring out how we track drought across such a large and diverse state as ours, and how we provide information that's useful for people making decisions directly on the ground. So, today's webinar will introduce a new set of drought indicators being developed specifically for Alaska that combines weather observations with computer models, and our speakers, Charlie and Rick, will talk about how these indicators are developed, some of their strengths and limitations, and how they could be used to support environmental monitoring and decision-making across Alaska. They'll also be looking for your feedback today. These tools are being developed, so hearing from people who may actually use the information will help make sure the final products are useful and meet the needs of communities, resource managers and others across the state. So we will be having a poll towards the middle or end of the presentation, so I'll let Charlie and Rick talk a little bit more about that. But before we get started, a few logistics. Everyone, you've all been muted, and your video has been turned off, and so, there will be a chance towards the end of the presentation, to ask questions directly, and at that point, I'll allow you to unmute yourself or turn on your video if you want to do that, but in any case, feel free to drop questions into the chat. Any questions we don't get to, we can bring up in the Q&A at the end. I know Charlie or Rick may be monitoring the chat and answering questions along the way as well. This presentation is being recorded and will be on the ACCAP website later today. I'll also send an email with the webinar link later today as well. Let's see, so, for those of you who don't know me, my name's Ed Plumb, and I'm the Weather and Flood Hazards Specialist here at ACCAP. And I am pleased to introduce our two speakers today. We have Charlie Parr and Rick Thoman. Charlie is a geospatial programmer and analyst with the Scenarios Network for Alaska and Arctic Planning. They're otherwise known as SNAP, here at the University of Alaska Fairbanks. Charlie develops workflows and web-based tools that transform climate and geophysical data into more accessible and useful resources for researchers, resource managers, and decision makers across Alaska. And many of you probably already know Rick Thoman, who is our climate specialist here at ACCAP. Rick has many years of experience producing and communicating climate information for Alaska. And his work really spans the bridge between climate science, Alaska communities, and the people who need that information. So together, Charlie and Rick have been working on these new drought indicators and tools, and today, they're going to walk us through what they've developed, and how these indicators work, and where they hope to take this work next. And a lot of that's going to be some of the feedback they get from you all on the call today. So, guys, thanks, pleased to have you, and I'm gonna pass it off to you. Rick Thoman All right, great. Thanks very much, Ed. And so, as Ed has mentioned, this is not going to be exactly like many of our webinars, where we leave some time at the end for, some questions. What you're gonna get today is really, a work in progress, and to move this work forward, we want to hear from you. So I'm gonna start off just with a quick review of what drought is, what it means in Alaska, traditional tools that we've had available for drought monitoring. Then Charlie's gonna introduce some of these new tools that we've been developing, take you through how they were developed, and show you a number of examples. Again, these are very draft. The main, most important part of this webinar is your feedback on that. And any way you want to provide that feedback, we will be very happy to get. We'll have time at the end where you can voice your comments or suggestions. Of course, you can put comments in the chat at any time. As Ed mentioned, we're going to try something new this go-round with a Zoom poll, and you can email Charlie and/or me directly with your comments, so whatever way works best for you. Alright, so, what is drought? So, when you think of drought, maybe you think of something like the picture there on the right-hand side, a dry, barren, agricultural area, like a cornfield in Kansas. On the left-hand side are a number of, kind of short definitions of drought. So, a drought, a shortage of water over an extended period of time. Okay, well, what's a shortage? What's an extended period of time? That's from the Weather Service, from the National Drought Mitigation Center, drought is less precipitation, rain or snow, over a period of time, defined here as a few months or longer. So, from the Drought Mitigation Center, say a month-long precipitation shortage wouldn't meet that definition of drought. And then, kind of, again, a fairly broad brush definition there. From Wikipedia. So, get the idea. Drought has something to do with a shortage of water or precipitation, maybe water supply. So that could be precipitation, or surface water, or groundwater. So something about low amount of water somewhere in the environment. Now, whatever definition you use, we can pretty easily divide up drought into different, varieties. So, for the precipitation version of the drought definition, the meteorological drought is most appropriate, so there is a prolonged period of time with less than average precipitation. So, again, some ambiguity there, what counts as prolonged, and do you just mean one hundredth of an inch below your normal of 100 inches? Would that be a drought? It's below average. Agricultural drought specifically, mentions the impacts on crop production or, livestock ranges. hydrological drought, incorporates some of that meteorological, impact, but also now explicitly, includes, things like mountain, snowpack. and its availability, and importantly, includes that locally significant threshold. And the photograph here on the, on the, right-hand side, during the, 2017-19 drought, this photograph of Wrangell at this point, when this photo was taken, was in severe drought, and that does not really look anything like the Kansas cornfield there. So, that local threshold is important. And then, the last one there, ecological drought, so prolonged and widespread deficit, that creates stresses across the ecosystem, so that could potentially incorporate some of the aspects of all the other definitions there. So different flavors, different ways of looking at drought. Now, so, the time over which precipitation is, or water supply is in a deficit or shortfall, has varied in these various definitions we've seen. Here's the graphic on kind of the timeline. So when we're talking about high precipitation, you know, in the space of hour, or a few hours, you know, you can have enough precipitation to cause problems. Drought is a little bit longer, lasting, but certainly in the week, or two week. Time frame, you can have flash drought when no precipitation accompanied by very high temperatures, and then as we move across the timeline to longer periods of time. You know, multiple weeks, month-type timescale, short-term drought, seasons, long-term drought, and of course, precipitation. can vary over years or decade, timescales. That might be, climate variability. So, in the, in the very long-term records in parts of the world, there is evidence of. episodes, eras, when precipitation was much higher or much lower than, say, the millennial scale average. And, of course, it can also be climate change related as well. So, we've got various definitions, and we've got this timeline from short-term, in the case of drought, you know, measured in a week or maybe two, on up to very long timescales to contend with. So, drought in Alaska overall does not look like drought in the Midwest. One of the big impacts that we saw in the 2017-19 drought in southeast Alaska, and to some extent in South Central as well. A hydropower and fresh water supply, particularly in southeast Alaska. It's a temperate rainforest. All of Southeast gets a lot of precipitation, so communities that rely on hydropower for their electric production, and drinking water. In general, they don't need lots of storage because it rains a lot. And that brings in the local, the local, climatology. What, infrastructure is built for. River transportation, of course, can be an issue with low precipitation, the vast areas of Alaska that rely on transportation, boats and barges up and down the rivers. can be impacted by low precipitation. Overland travel, of course, if you're expecting snow on the ground and there isn't any, like this picture here from February of 2025, Iditarod was moved because there was a large stretch on the north side of the Alaska Range. That had no snow on the ground. This was an example. of a snow drought. This area had snow on the ground, big atmospheric river in January, washed away the snow, and then, it didn't precipitate again. Wildfire is a big issue for drought. In Alaska, the boreal forest, the fuels dry very quickly. When it's warm and dry, this is our, this is a primary issue, in the summer, of course, and, If it's really warm and dry, the fuels can dry out really amazingly quickly and be susceptible to lightning starts or human starts. Of course, as you might expect, agriculture and gardening can be affected. And one that people don't think about very often, but it has caused significant problems. If we have low precipitation, in particular low snow, in the autumn and it gets really cold, that the ground will freeze to significant depth fairly quickly. About 10 years ago, this was a problem in the Palmer and Wasilla areas, where there was no snow on the ground into November, into December, but cold, low temperatures, and so we wound up with water supply that was buried, was freezing, and that also happens even In the interior, typically those frozen water supply pipes wind up freezing up later in the winter, but definitely an impact of low precipitation or low snow. Water, river temperatures are also affected by low precipitation. For salmon and other anadromous fish, they like water temperatures not to be too warm to return to, and that has been a problem in the past. Forest health as well. Some insects are more common in dry summers, and we saw examples of drought-related tree damage in southeast. during the 2010s drought. And, as one that many Alaskans are familiar with, berry production. Very low precipitation, and you wind up with a whole bunch of shriveled berries. So, our drought concerns here in Alaska are a bit different than the lower 48, but we definitely are impacted when we don't have the water that, the environment and we are expecting. Now, when you think about monitoring precipitation, you probably first think of weather station observations, and that's what the pictures are here, all individual, different kinds of weather observations that are available within Alaska. The problem that we have using point observations is that the number of observations, especially, but not limited to rural Alaska, is quite limited, even in more densely populated areas, because Alaska is the land of complex terrain. There are very few observations available at elevation, and of course, that water counts, too. The Natural Resources Conservation Service, for decades has measured the water content of the snowpack. That's mostly done, during the end of winter and spring season. There are, quite a number of automated observations that do that, but they are concentrated in South Central, and to a much lesser extent, in the interior. There's only a relatively few of those in southeast and, Western Alaska. Another traditional measure of drought, it would be to monitor stream flow, the amount of water in rivers. In Alaska, that is much less useful than down south, because of High elevation snow and glacier melt, so rivers that are fed by glaciers or long-lived snow areas can run high even if there's no precipitation. Historically, soil moisture has not been much of an issue in Alaska, mostly because it is measured almost nowhere, and we have no real mechanism to monitor that. That potentially is changing. Another traditional measure is vegetation growth or damage, as we saw in the Kansas cornfield picture. However, in Alaska. In most of the state, not all southeast, at least low elevation, somewhat of an exception, vegetation growth in the summer is more limited, by temperatures than precipitation, so that is not as useful here. We also have, besides the point-based observation, we also have things like the gridded, so maps without gaps type estimates of precipitation. The graphic here shows an estimate of the January 2021 precipitation. from, National Centers for Environmental Information, This is, this is basically takes the available point observations, but then, smooths them out across the, in our case, all of the state. Again, it is limited by what observations are available, projected onto some kind of, climatology as a background field. In the lower 48, a highly valuable source of precipitation information is the PRISM database. That is run out of the out of Oregon State University. However, at the daily scale, that is not available in Alaska, so that is out. How about remote sensing? Radar. The vast majority of Alaska has no usable radar coverage. There are, there's no weather radars north of Fairbanks. Some of the radars are severely terrain blocked, Biorka Island near Sitka. Nome, the northern half of the radar is beaming into the mountains to the north. And even where there's better, again, because of the small coverage, it's not really useful at the state scale, and it's really best at the individual storm-type scale. Satellite estimates are widely used in low and mid latitudes. for, again, kind of event scale, storm scale estimates of precipitation. The still quite limited usefulness of those satellite estimates of precipitation at high latitudes. There's a lot of work going on, to try to improve that. But at the moment, it's still pretty limited. And that brings us to, our latest tool in the toolbox, and that's our climate analysis model, that, we will be spending much more, most of the rest of this presentation, on. So, what are climate analysis models? So, as Ed mentioned, it's a blend of observations with short-term 12-hour or less weather forecasts using modern weather models. The jargon term for this is reanalysis. So, reanalysis is just think of it as climate analysis models, and we get these map without gaps. Example there on your right-hand side, this is as ex-Typhoon HaLong was approaching, southwest Alaska last October. This is from the NASA MERRA reanalysis product. Now, reanalysis is not new, it's been around. Since the mid-1990s, but it's only in the past, really in the 2020s, that it's been available at a time scale and a resolution Horizontal resolution that we can make use of it in near real-time environmental monitoring. And by far the most widely used reanalysis around the world right now for near-real-time monitoring is the ERA5. That's a much more concise way to say the European Centre for Medium-Range Weather Forecasts Reanalysis Version 5. again, widely used for many purposes around the world. It's currently the state of the art. In part because of the high resolution, in part is available, at very, it's available, within about 5 days of the current. So that is about the, that's the best we have right now. The, these, like, the MERRA example here, that's only updated at a monthly time scale. So, ERA5 comes in a number of flavors, ERA5, the basic. The underlying atmospheric weather model is from 2016. Grid size of about a little over 300 square miles, that might sound big. That is amazingly fine resolution compared to what was available. Even a decade ago. Because it's a weather model, a vast array of meteorological variables are available. However, ERA5 imports ocean temperatures and sea ice from other data sets. ERA5 is available since 1940, so even before the start of upper air observations. Or systematic upper air observations, to about 5 days before the present, and available as fine a timescale as hourly, and you can aggregate, of course, up from there. Now, the flavor that we will be looking at today It's called the ERA5-Land. Now, importantly, it uses the ERA5 atmospheric Model. It's available since 1950, but it has a different land, model that is part of that. So, the grid size is much smaller, about 35 square miles on a grid, so that's quite, that's quite fine. However. remember that the underlying weather model is not that fine. It's still using that ERA5 basic model from 2016. So, for parameters that are specific to the ground, we see significant improvement. For things that are really mostly or entirely atmospheric-driven, like precipitation, the difference between ERA5 and ERA5-Land is really, typically quite small. Here's an example of the difference. This is an example of the April 1st, 2025. This was the percent of median snowpack, so the amount of water in the snowpack. Very important parameter for Alaska in the spring. And, on the left-hand side, you have the ERA5, so this is the base, the relatively coarse model. And on the right is the ERA5-Land, and I've circled areas there on the ERA5 base, where you can see there are some substantial differences between the two. In some areas, it's in pretty good agreement. For instance, southwest Alaska, Bristol Bay, both show, very low snowpack, and both show, well above normal snowpack on April 1st, 2025, across the central interior. But big differences there across the Seward Peninsula. I'm not sure what's going on there in the Yukon there between the two. It's the same in the northern Yukon Territory. Also, in the Anchorage area, you can see a difference in the resolution. Of the two models here, the, a big difference there between the western Kenai Peninsula, way below normal, and then in the mountains of the Kenai Peninsula, showing an area at least somewhat above normal. Work has been done on verifying these two. Here at ACCAP, we had a Hollings scholar a few years ago looked at this in depth in river drainages that had substantial observations, and ERA5-Land is much better. Okay, how can this help us with, drought monitoring in Alaska? Okay, it's available statewide, as you've seen, includes things that we have some, point-based observations, like precipitation, but it also includes things that there are effectively no measurements of. Things like soil moisture, things like evaporation. ERA5 goes back long enough that we have available the data to derive multi-decade statistics, so averages. So we can do things like percent of median on any given day, or month, or even hour if we wanted to. And because of this data availability, it's really easy to calculate what becomes standard drought monitoring indexes, like the Standardized Precipitation Index and the Standardized Precipitation Evap, Precipitation Evapotranspiration Index. Charlie's going to go into more detail on that. Caveat. The ERA5 looks really detailed, and some ways it is, but the overlying atmospheric model is still relatively coarse. And as high a resolution as ERA5-Land, it is not adequate. It is not, the area covered in each grid box is too large to really resolve what's happening in the complex terrain, really complex terrain. That includes most of southeast Alaska, includes the Gulf of Alaska, the Alaska Range, Cook Inlet area, as good as ERA5-Land is, it's not yet, we don't have a operational climate analysis model that is able to work with those very complex terrains, big elevation changes in very short distances. That is what I have for you, and I'm going to turn it over to Charlie. Charlie Parr Thanks, Rick. Alright, I'm Charlie Parr, and I'm with the Scenarios Network for Alaska and Arctic Planning, maybe better known as SNAP. part of the International Arctic Research Center at the University of Alaska Fairbanks. So today, I'm going to take you through some of the methods that we use to turn that climate reanalysis, those maps without gaps, into drought indicator products like you're seeing on the right. And we'll unpack what that figure is showing you there. But I'll take you through some of the methods, show you some results, and then once we're all on the same page. what I'd like to do is have some time to ask you a couple specific questions about how we can zero in on really the most, important, slices of this work, slices of this data for whatever your application may be. Alright, next slide, please. One back. I think we might have double tapped. Perfect, there we go. So what did we actually do? Well, we built an automated pipeline that generates a multi-scale drought monitoring suite for Alaska. So, on the left, we list the five drought indicators that were computing, and these account for a variety of climate factors. Precipitation, snowpack, evaporative demand, and on the right, we're looking at each one of these indicators over 7 different, retrospective summary intervals. So, 7-day all the way through 365 days, because drought evolves over multiple time periods. So this 5x7 matrix, gives us a few different lenses, for looking at drought conditions, you know, in Alaska in the context of ERA5-Land reanalysis. Next slide, please. Alright, here's a diagram slide. Sorry about that. We'll try to move through these quickly. So, how are we doing this? Well, we're downloading ERA5-Land reanalysis data from the climate data store, and we're getting, the variables that you see in the center box there. So, we have, Daily total precipitation, we have daily total potential, evaporation, we have the daily average snow water equivalent, and then we have Daily volumetric soil water content, for layers 1 and 2, which together represent about, the surface to 28 centimeters depth. below the surface. And this data download happens sort of in two pathways. There's one for the 1981 to 2020 baseline reference period. We just download that data once. This is a really important thing to keep in mind here, is everything that we're going to look at is in comparison to this 1981-2020 40-year baseline reference period. And then with each, recurring, you know, computation of these drought indicators, we download the most recent year's worth of data. And it's also important to keep in mind that everything that you're gonna see, the best available or most recent data that we can get is always going to be about 6 days, behind, the present date, basically, so it's not real-time data. Next slide, please. Okay, diagram slide 2 of 3. This is by far the worst slide of the webinar, so just hang in there, everyone. The heavy computational processing here is, churning through all 40 years' worth of daily frequency data for each of those five variables. So there's sort of two tracks here. On the right, we're constructing day-of-year climatological normals. So for soil moisture, we're combining the two ERA5-Land layers using a depth-weighted average. For precipitation, snow water equivalent, and soil moistures, you know, we group all 40 years' worth of daily values, and compute long-term daily means. And so those are the reference day-of-year climatology or baseline normal values that we're going to compare what's happening now against. On the left, we're calibrating our standardized drought indices, the, Standardized Precipitation Index, SPI, and Standardized Precipitation Evapotranspiration Index, SPEI. We'll talk a little bit more about those, but for those, we compute some rolling precipitation or, water budget totals, and then we fit, probability distribution to, every grid cell and every day a year. And then we have have those parameters to compare the recent conditions to. Okay, next slide. Alright, so, last diagram here. All that's left, then, is to compare the recent data to the baseline data, and, For each summary interval, there's sort of two pathways here. We do the recent versus the climatology comparisons. So, for precipitation, we sum the recent precipitation, divide by the climatological sum over that matching day of year window to get the percent of normal, and then similar for snow water equivalent and soil moisture, except we're Looking at the mean conditions for that window. On the right, we compute the standardized drought indices, SPI and SPEI. So, instead of dividing by a recent instead of dividing the recent data by the normal data, there's sort of a probability transformation happening here. We take, for example, a 30-day rolling precipitation for SPI, and we say, okay, you know, what are the pre-computed distribution parameters for that same period? And then we use a cumulative distribution function to to determine the probability, and then we convert that into a standardized normal score. So this is a pretty conventional implementation of SPI and SPEI. Next slide, please. Okay, great. So, the maps you're seeing now, all represent an analysis date of August 13th. So what you're seeing here is the 30-day retrospective summary interval for all five drought indicators. So the reference date here is August 13th, so this is sort of reflecting, like, a mid-July to mid-August window here for 2026, for this year. Across the top, from left to right, we have precipitation percent of normal, snow water equivalent percent of normal, SPI, the standardized precipitation index, and then on the bottom row, we have SPEI, the standardized precipitation Evapotranspiration Index, and then soil moisture percent of normal. We'll take a look, a closer look at each one of these indicators, but we can already see there's some interesting tilts, across the state here, and then we also see some snow there in the Brooks Range. It snowed at Tulik Field Station on July 27th, so the reanalysis is picking that up as well. Next slide, please. Alright, soil moisture, percent of normal, our first drought indicator. This is a zoomed-in view of southwest Alaska, and the summary intervals are increasing in time, left to right, top to bottom. So, 7-day, 14-day, 30-day across the top, 60, 90, 180, and then 365 at the bottom. This, view doesn't have the most extreme conditions present, I would say, except for maybe the 20-30% soil moisture deficit in the top right 30-day view. There's that orange patch, kind of near Aniak and Mountain Village. But we do see an evolution in time here. At the 365, the one-year look-back period, things are moist. But if we consider the last two months, maybe that 60-day panel, the 30-day panel, there's a modest, but, you know, pretty widespread soil moisture deficit. Next slide, please. Okay, snow water equivalent, zoomed into interior Alaska. As we might anticipate, in mid-August, the shorter look-back periods are not that useful, because, you know, the climatological normal condition is that there's no snow on the ground. You can see this in the gray areas, that's basically no data. One interesting thing here, though, is you can see, one of the weaknesses of a percent of normal approach, if the normal amount is some very small value, because it happened to snow a few days somewhere in that 40-year window, and the recent amount is zero. You know, you get 0%, and it can make things look pretty dry here. I would point out, though, especially in that, second row, center panel, there is a sort of interesting deviation that looks, to my eyes, driven by elevation, maybe less snow than what is normal for that 90-day period, prior to August 13th at the fringes of sort of the high elevation terrain there. Okay, next slide, please. All right. Hello, Southeast Alaska, how are you doing? If you're there in the chat, maybe you can chime in and let us know if you feel like it's been a nice summer. Because according to the ERA5 reanalysis, it has not rained nearly enough on you all down there. The retrospective summary intervals at, you know, 7 days through 60 days look far drier than what is normal. Although, as Rick mentioned, I do want to point out that one caveat, you know, with this climate reanalysis is that it can be difficult to resolve conditions in, you know, really steep topography, and southeast Alaska, of course, is a good candidate for that description. But I did read in the news, alright, it's sunny right now in Juneau. The data's been validated. Thanks, Allison. But I did read in the news that, you know, Juneau is having a top 5 dry, summer season so far. But if you sort of zoom out in time, and we look at the last year, the picture is, actually just about normal there. You know, it's all white in that bottom box. Okay, next slide, please. All right, so on to SPI and SPEI. What are these? So these are very conventional drought indicators that have a lot of usage in the United States outside of Alaska. And these don't have physical units. They are standardized statistical scores, or z-scores, that are centered at zero. So, zero is average. If it's above zero, it's wetter than average. If it's less than zero. It's drier than average, and the negative z-scores, the negative values map to these drought conditions that you're gonna see on the left. So these are from the U.S. Drought Monitor, D0 through D4, these map to values of SPI and SPEI, and these are kind of a conventional way to look at drought. And what these do is they allow us to Make comparisons across very different climates, so we can compare Drought conditions in southeast Alaska to the North Slope, and they can also, allow us to compare, or sort of smooth out effects of, like, wet seasonality versus dry seasonality, and they really let us answer how unusual is this precipitation deficit or, water budget deficit in the case of SPEI, that we're seeing. Because, you know, a 50% precipitation deficit in some places might be more common than in others. So this allows us to really just sort of understand how usual, or unusual, a condition that we're seeing might be. Next slide, please. Okay, so here's the standardized precipitation index. And we're gonna stay with Southeast Alaska here because we were just looking at precipitation percent of normal, and here you're gonna see that the 30-day indicator, so the top right panel. For SPI, it has a pretty widespread D4, D3, so exceptional to, I think, severe drought, area. You know, the 7-day picture, maybe the most recent week prior to August 13th, doesn't look as dire, but especially, in the 60-day view also, the area just south of Haines. That looks pretty alarming, to my eyes. Okay, let's go to the next slide. And now we bring in the atmospheric evaporative demand into the picture. So the pattern is changing again when we go from SPI to SPEI. The colors and units, are the same. So, you know, the darkest red is D4, the other bright red is D3. The pattern does change, and we're seeing maybe an even stronger drought signal. Especially over the longer look-back periods in that middle row, particularly maybe in the area of Glacier Bay National Park, sort of, west of Juneau, in the, in the images. So, those are all five drought indicators. Let's go ahead and move on to the next slide. Okay, so I would like to point out that the U.S. Drought Monitor, does have data for Alaska, and we do see a little bit of the same signal here. But I'm showing this because I think this is revealing some of the motivation for our work here. We do see some of those same, dry conditions emerging in this figure. And this is about the same analysis date. This is August 18th. We've been looking at August 13th, so it's a little bit different, not quite apples to apples, but we do see the yellow patches, there in sort of the Upper Copper River, Upper Tanana River Valley, and Southeast Alaska. Next slide, please. But I think this tells a much different story. This is the SPEI 30-day look-back period. We're looking at the entire spatial domain of our analysis here, which extends out into the Yukon, and I want to draw your attention to the, magenta circle there of the Upper Tanana River Valley, Upper Copper River Valley. And so, we've got a very strong, D4, D3 sort of anomaly showing up in the data, and I thought that this was interesting. this, data represents, you know, the snapshot of the 30 days prior to August 13th. If you've been sort of watching the news in Alaska, August 17th, the Mukluk Fire ignited, near Tok, which is not labeled on this map, but it's right there, kind of in that magenta circle. And I think that this picture Compared to the previous slide, tells a much different story. So that's sort of why we're here. Next slide, please. Okay, now for my favorite part, of today, I think, is, we need feedback from you to understand which slices of this data are the most impactful. There's a lot of different indicators, there's a lot of different summary intervals. You know, we're showing 7 here. And I want to know what decisions could this information help you make? Maybe something that you can't do today, but if there was a perfect version of this data, you could make that decision, or take that action if it existed. And so, in the chat, or if we've got the poll up, if you could keep, you know, maybe just 1, 2, 3 of these 35 views, what do you choose, and what would you use them for? I'm just gonna pause right there, for a little while. Edward Plumb Yeah, and Charlie, I can pull the poll up. I actually had it polling for the last handful of slides, but I do have, I do have a poll question, one of the questions applies just to this slide specifically, so, I didn't know if people wanted you to kind of scroll back and review the different slides, or I'll bring the pull up anyway, and we'll see how this works. Rick Thoman You want me to go back, Charlie? Charlie Parr Sure, let's go back, To, maybe the, yeah, the full domain slide there. Edward Plumb So, can you guys see, can you guys see the poll? Hey, Charlie. Oh, Allison, yes. So, the way the poll is set up, yeah, full domain, yeah, which slide is that one? Charlie Parr Yeah, maybe it's, 19 or something like that, just the first, Edward Plumb Yeah, slide 19. So this is the, yeah, slide 19's the first one you showed with your graphics, and so, The way the poll's set up here is, each question has, you can comment on slide 19, so you can see in the poll there's a, there's a image of slide 19 there as well, so if you want to add any comments to that slide in particular, and I don't know if, Charlie, you have anything to say about that slide again, now that you've kind of, everyone's seen this whole suite of slides that you've presented. Charlie Parr Yeah, I guess just to review, these are the five drought indicators that we're currently producing. What you're seeing here is the 30-day retrospective summary interval, so the month prior to August 13, 2026. That's what's being represented by this data, and everything is compared to what the 1981 to 2020 normal would be, for that same part of the year, so that same mid-July to mid-August period. Left to right, precipitation percent of normal, snow water equivalent percent of normal, SPI, and then on the bottom, left to right, SPEI, and soil moisture percent of normal, Hotter colors indicate drier, SPI and SPEI, you know, share the same D1, D2, D3, D4, categorization, and the rest are expressed as, you know, percent of normal. And, so that's sort of, I guess, a quick summary, I guess, of everything that we saw, just from a zoomed-out view. Edward Plumb Okay, thanks, Charlie, and people should be able to fill in just free form into that, to comment on that particular slide. If you want to, there's a couple questions in the chat, I don't know if you want to address those while people are kind of looking at that slide. Yeah, the, Rick, this question may be for you, though, or Charlie, you may know. The first question was, does aquifer depletion qualify as a form of drought? Rick Thoman I would, so, in, in, the flavors of drought there. Groundwater supply was, in some of that, so, Aquifer, water supply? I'd say, sure. Edward Plumb And then, we had one more question, this was referring to the, Mukluk Fire. I failed to understand how the fire near Tok influenced the SPEI. Should it not be opposite of what you presented? Charlie Parr Yeah, so I may have misspoke, I'm not sure. So, the SPEI, which is the bottom left panel in the figure that we're looking at, is showing very dry conditions driven by a lack of precipitation and enhanced, atmospheric evaporative demand. It's showing a very unusual combination of those two factors, for the Tok area, for the sort of month leading up to the ignition of the fire near Tok. So, perhaps I, when I was speaking, I flipped the cause and effect there, and of course, you know, I would hedge that I'm not a fire weather expert, but I think we are seeing, you know, probably a real signal there in the SPEI, at least to some degree. Rick Thoman And I would just add, in the SPI, so the upper, right-hand graphic there, which is just precipitation, you also have those darkest colors, so that D4 exceptional drought in the upper Tanana Valley there. Edward Plumb Yeah, and Allison had a remark there that, the strong, dry conditions were noted in Tok there. Charlie, I don't know if you want to go to the next slide for people to comment on? And so, for those of you, this is number 2, slide 20 comments, and so you can see that in your poll, if you have any comments on, on that slide, or the way that data is presented. Charlie Parr Yeah, and I guess one thing to note here is that, we did truncate the top end of the color bar here, just to sort of emphasize what areas might be in deficit. So everything greater than zero, or everything greater than normal is green, I guess. So that's worth pointing out here, is that, you know, you might have more variation if we actually looked at, okay, what's, you know. 20% wetter than the normal value for that same period, would, I guess, be one other thing I would, I would add to this. Edward Plumb And, yeah, whenever you're ready to go to the next slide. Charlie Parr Yeah, next slide. Edward Plumb 2 again, Charlie. This is, number 3, slide 21, comments. Charlie Parr Yeah, so snow water equivalent percent of normal. It would be interesting to take a look at this one. I think in the spring, certainly we'd have, perhaps a more interesting, signal there during the melt season for those shorter summer summary intervals. And, I think when I went through this slide the first time, I just underscored, you know, percent of normal can be a difficult metric to work with when the normal value is very small or zero. And the other thing to point out about to point out in ERA5-Land is that the snow water equivalent value is capped at a very large value, and that is to reflect, basically glaciers, or ice caps, and that's what you're seeing in the Alaska range there. Those values are pretty much always white, so they're pretty much always a hundred, you know, right there at their normal value, because they're Sort of indicated as glaciers or ice caps in that, In the reanalysis product. Edward Plumb And I failed to mention, too, and your comments are anonymous as well, everyone. Charlie Parr Yeah, so, you know, really let us hear it. We can go to the next slide. Yeah, so this, Edward Plumb This is, number 4, slide 22, comments. Charlie Parr yeah, precipitation percent of normal, zoomed in on southeast Alaska, I don't know. Yeah, I guess, Rick, if there's anything here you want to annotate in terms of, you know, geographic patterns, But I think, yeah, the picture is certainly the last 6 months, you know, precip deficit for Southeast Alaska writ large, although perhaps still Around normal for the year. Rick Thoman Yeah, just, as you mentioned, Charlie, as I mentioned, just to reiterate, can't say it enough. Again, because of the, of the very complex terrain in Southeast, all the state-of-the-art reanalysis is really not what you're gonna get if we had a thousand observations in Southeast at all elevations. So, keep that in mind. More so than, more so than, say, portions of the interior or north slope. where the terrain is gentler. It's a good example of how this reanalysis at this stage can be a tool, but it's not definitive. Charlie Parr Sure, let's go to the next slide. Keep things moving. Yeah, I think we'll stay with the maps here. So, here's our standardized precipitation index, same southeast Alaska overview, and you can think of the numeric values, on the y-axis, they might be quite small here, but for example, the deepest red color, that D4 category at the bottom of the color bar, the numeric value that corresponds to that is less than or equal to negative 2. And so you can think of that as being Greater than two standard deviations. away from the normal condition, because it's a standardized z-score. And so that helps us understand, okay, the precipitation deficits we saw in the previous slide, how unusual is that versus just it's a 50% deficit? Edward Plumb And this is, number 6, slide 24, if you have in the, in the poll for comments, if you want to put that there. Charlie Parr Thanks, Ed. Rick Thoman And I would, I would just point out, on the percent of normal precipitation, like at the really short time frame, like 7 days, you know, very low precipitation. But in any given day, 7-day period, in the early to mid-August. in northern southeast, that's not that uncommon. And so, say, around the Haines area, which was, you know, quite deep red there on just the percent of normal. This helps, the SPI helps to, put that in context. A very dry 7-day period, not that unusual, say, in northernmost southeast. Charlie Parr There's a question in the chat about, soil moisture layers. I'll read it aloud here. Does the definition of soil and the two soil moisture layers include the flammable organic material parentheses, duff layer, or just the mineral soil. That's a good question. I might have to defer to Rick on this one. My understanding is that there's soil moisture, the way the soil moisture layers are prescribed in ERA5 is there's the, top layer, which is layer 1, 0 to 7 centimeters below the surface, so I'm guessing that that Would probably include, that, organic material layer, and then there's layer 2, which is 7 to 28 centimeters, below the surface. But I don't know the full picture of, the ERA5 soil moisture physics, so, I would have to follow up on that. Edward Plumb And guys, we're about 4 minutes from the top of the hour, and I know. Charlie Parr Oh gosh, okay. Edward Plumb We have another slide. We can, we can send this out in the email with the recording, this poll, like, as a Google survey, if people want to add, more comments and have more time to look at it. So, that's an option. Charlie Parr Great. If there's not much time left, maybe we can just go through to the, I think I had another question up, a couple more slides. Yeah, so, you know, the other thing that I wanted to ask people was just, what would you need to see before you could trust this information. You know, what are sort of the hurdles we'd have to clear? And then, really, what's coming next? well, we're here to listen to you and zero in on the most, you know, important parts of this data. We can modify the summary intervals. If you don't like 14-day, you really need 20-day, like, that's something we can do. We can change the data viz, we can change other ways to explore and present this data, But probably fundamental changes to, like, the spatial domain, or, like, the input variables, or, like, an entirely new drought indicator. are not on the table, you know, right now, you know, pending, like, an extension of this work through additional funding. Rick Thoman So yeah, again, appreciate all the feedback so far, and we'd like to get more. Any way you want to provide that to us, we would be very happy, to get it. We'll be capturing the comments and questions here in the Zoom chat and poll, and if you want to say something, now's your chance, or email us. Edward Plumb Okay, and I see Aaron's got his hand up. You can unmute yourself, turn on your video. If you want to. Aaron Jacobs Great, thanks, Ed. Can you hear me okay? Edward Plumb Yeah, hear you loud and clear. Aaron Jacobs Great, thanks. Yeah, thanks for this work, it looks really great, and I'll probably talk to you guys offline, talking about different durations and timeframes for these SPIs, and really do like those statistics as a really good indicator, like, focusing on putting precipitation, deficits and surpluses in the context. But duration is definitely a certain context around here. 3 months, 4 months, to a year, and so, I'll talk to you guys offline about it, but, really great work on this stuff. Appreciate it. Rick Thoman Thanks. Edward Plumb Thanks, Aaron. We've probably, we've got another time for one more question, if anybody wants to unmute themselves and ask a question. Okay, hearing none, I think we got to all the questions that were in the chat, and like I said, I will, I'll send out an email with a form if you want to have a closer look at those slides that were in the poll to add more comments, and then we'll also include Rick and Charlie's email to get directly in touch with them if you have a question. Rick Thoman Yeah, thanks a lot, everyone. We do want to acknowledge that this work, was funded by, the National Weather Service as an effort to, improve drought monitoring in Alaska, and, we are grateful for that support, and, hopefully we can get these, the ultimate goal here is, get these products tweaked based on, feedback, and then, get them, on a website online, where they will be updated, every day, with that lag that Charlie mentioned, but, that this would be updated on a daily basis and, available for all kinds of, low precipitation type monitoring. Charlie Parr Yeah, thanks everyone. Really appreciate being part of the webinar, and yeah, please do get in touch with me and Rick, if you'd like to help us continue this work. Edward Plumb Great, thank you guys, and Appreciate everybody joining us today, and you can check out our upcoming webinars in September on our webpage. We've got, starting to fill up our fall lineup of webinars, so we look forward to seeing some of you again. Okay, have a good, have a good day. Rick Thoman Thanks.