[00:00:22] Bridget Scanlon: I'm pleased to welcome John Nielsen Gammon to the podcast. John has been the state climatologist of Texas since 2000 and is also serving as a Regents Professor of Atmospheric Sciences at Texas A&M University, also Director of the Southern Regional Climate Center. So multiple positions, not sure how you do it all John.
And he has received numerous awards during his career. And today I would like to emphasize the award you received from Sigma XI for being an excellent communicator, and I'm sure we will benefit from that today. So thanks so much, John, for joining me.
[00:00:58] John Nielsen-Gammon: All right. Yeah, you're welcome. My pleasure.
[00:01:01] Bridget Scanlon: You've done a lot of different types of research on climate related issues John, and today I think we're going to focus on extreme events and talk about your recent presentation to the legislature after the central Texas flooding, which was excellent. And then go to the other extreme, talk about drought. And then also the report that you developed for Texas 2036 on climate extremes.
I think on top of everybody's mind is the recent Central Texas flooding with so many lives lost about 117 people. And I thought your presentation to the legislator was very informative I think you mentioned 10 to 15 inches of rain in Kerr County resulting in flooding in the Guadalupe River.
And I guess the intensity was three to four inches per hour. And you indicated that you didn't think it was possible to predict the location of the flooding, one to two days in advance. And also that the weather warnings could not keep pace with the threat and change and action necessary with such rapid rises of flood stages in the river.
So maybe you can describe a little bit more about the situation during July 4th flooding in the Guadalupe.
[00:02:14] John Nielsen-Gammon: Sure, yeah. The setup that you need for very heavy rainfall in Texas is essentially, sounds obvious, but lots of moisture. But that means not just moisture at low levels, but also moisture higher up because if you've got a thunderstorm developing the dry air in the atmosphere will choke it off and cause rain to evaporate before it reaches the ground.
So in this particular circumstance, it's the summertime, so we've almost always got plenty of moisture at low levels. But in the higher levels, we're talking like a mile up, two miles, even five miles. We had moisture coming from northern Mexico, which had gotten there through some combination of tropical storm in the Atlantic and a hurricane in the Pacific.
And getting that feed of moisture allowed a situation where you can get extreme rainfall. Of course, we had extreme rainfall watch in place, but it covered lots of Western and Central Texas because there was a lot of moisture in that whole area. But what finally ended up happening, unfortunately, was in a few locations, thunderstorms developed and were stationary.
They didn't really move. For that to happen, you've gotta have differences in the wind at different levels in the atmosphere that can hold the storm in place while you're still getting a feed of moisture coming at low levels. So that happened in northern part of San Angelo. It happened in Mason County.
It happened in Travis and Burnet County the next day. But the flood that killed most people was the flood in Kerr County where thunderstorm just developed and grew and stayed in place for several hours. So we had a sense that heavy rainfall was possible. The forecast really didn't mention the possibility of something more than seven or eight inches.
So that was a shortcoming in the medium range forecast, but then you can't really warn on the storm location until the storm actually exists and you see where that particular type of storm has developed. When that happened, they got the warning out fairly promptly, a little after one in the morning.
But unfortunately, the storm just sort of stayed there. And the updates had a hard time keeping up with how much rain was actually falling. And I think part of the problem was the Weather Service didn't have a direct connection to the actual amount of water in the South Fork of Guadalupe until it finally reached Hunt and Ingram, and moved down toward Kerrville where the gauges were located.
[00:04:49] Bridget Scanlon: Yeah. And of course being a holiday and then the storm starting in the middle of the night and all of those things were difficult to deal with but also, maybe some buildings in the floodplain and many different aspects. I thought it was interesting in your presentation to the legislature, you described many of the similar floods in the past to the recent flooding, and I thought that was very helpful to understand. This is not the first time.
When I was reviewing the various ones you mentioned, it's interesting how many of them occur on holidays. And of course, if they happened in the middle of nowhere and there was nobody there, it wouldn't be a problem. So I thought it was nice to review the storms in the past and the flooding issues.
So that was a good context I felt.
[00:05:36] John Nielsen-Gammon: Yeah, every flood is unique. So there are several different storms that have the high water marks at different points along the upper Guadalupe. Probably the one that's most similar was back in 1935 which hit the South Fork pretty hard. And there were summer camps along the Guadalupe back then, but there was no loss of life with the flood. And part of the reason for that was it happened in the daytime, where people could see the water rising. It was easy to recognize the hazard and get out of the way of the water. When something happens at night, you've got a lot less sense of what's going on. The heavy rain and lightning is very intimidating, so you're reluctant to get outdoors.
And so that was unfortunately one of the dominoes that fell that led to the catastrophe that we experienced. But even at that there's 10 to 15 inches of rain there, but that's, that pales in comparison to some of the other rainfall events that we've seen historically in the region.
The one that actually is record setting on a world scale was over 21 inches of rain in less than three hours. So that's almost twice the rainfall rate that caused the Kerr County flood. So the flood waters literally could have been twice as large if that sort of thing had happened in the wrong place, and again, at the wrong time.
[00:06:58] Bridget Scanlon: And was that the, the one in D'Hanis? Was that...
[00:07:01] John Nielsen-Gammon: Yep. Yep, that's right. Yeah. The rainfall was measured at a ranch about 10 miles up stream of D'Hanis. And by the time the flood waters got to D'Hanis, they were so big that actually overflowed into the next creek over.
[00:07:12] Bridget Scanlon: Yeah. Some of the others that you mentioned, the 1921 storm in Thrall, Texas, in September. 36 inches of rain in 18 hours. And that impacted Austin and Thrall. And that was a US record, I think.
[00:07:28] John Nielsen-Gammon: For a while there. And the same event, flooding in San Antonio and basically the reason or motivation, at least for the implementation of the flood control system that's in place now which includes the Riverwalk, which is probably the second most famous aspect of San Antonio these days behind the Alamo, which we all remember.
[00:07:48] Bridget Scanlon: Yeah. We are no stranger to these extreme events. And I guess the state is doing more and more on flooding. Texas Water Development Board has the flood planning efforts and trying to map the hundred year, 500 year floodplains, and do a lot of modeling analysis and bottom up approach.
So it's nice to see that, and you are currently doing a study for the Texas Water Development Board, collaborating with Rice and other groups, Rice University, on extreme events and rainfall return periods. Can you describe a little bit, the methodology used to do that type of work?
And I know you mentioned that the work should be coming out in the Fall. Your publication should be coming out this Fall.
[00:08:29] John Nielsen-Gammon: Yeah, we, it's been about a almost a three year project at this point. And the basic idea is that we have seen across most of the United States, including Texas and the Gulf Coast region, an increase in extreme rainfall. And most of that, possibly all of it, is due to the rising temperatures that we've also experienced, both locally and in the Gulf of Mexico and globally.
That leads to warmer temperatures in the atmosphere, which increases the capacity, the atmosphere to hold water. And that means that a given storm system will end up producing more rainfall at a greater rate. Of course, you can also get changes in, in where the storms exist and changes in storm structure with a different climate, but that seems to be the temperature thing seems to be the main factor driving the increase across the southern United States. So the federal governments looking at that sort of stop, start with the Trump administration, but they've made a lot of progress in analyzing trends in rainfall, extremes and looking at climate model projections.
But we are following a somewhat different approach than they are. They're treating the trend as the historical trend and whatever that is locally. That's what's been happening historically. And then they patch onto that the projected trend from climate models.
A couple problems with that. First off, you can't really trust local trends. I talk about rainfall extremes increasing overall. You can only really tell that if you look at trends over a fairly large area because the randomness of the weather we call natural variability can affect a trend locally. There's a big difference in the extreme rainfall trend between say Houston and Victoria because Houston's had Harvey and Allison and Memorial Day flood, et cetera, et cetera.
But there's nothing special about the climate of Houston compared to Victoria that would make things happen in one place but not the other. So local trends we don't think are as reliable as large scale trends.
Second difference, the historical trends tell us something about the future trends. Models aren't perfect, so looking at what's happened historically gives us a clue as to what to expect going forward. And likewise, even though the models aren't perfect, you can run lots of models and average out natural variability. So they help to tell you what the climate change signal might have been historically.
So rather than looking at those two periods separately and those two information sources separately, we're merging that information, and getting what we consider the best estimate of both the historic and future trends that are consistent with each other, and consistent with the other estimates of extreme rainfall that we're producing from the data.
So those are the main differences. We're finalizing that merger process. We expect to see some numbers preliminarily within a month. And, once we trust them we'll deliver to the Water Development Board and they'll be available for use for people who want to take that into account.
An important thing to remember is it matters not just for the future planning, but it also affects the present. If extreme rainfall, for example, is increased by 10% over the historical period, if you don't take that into account, then your estimate of extreme rainfall at the beginning will be 5% higher than what it really was, and at the end of the period, the present day, it'll be 5% lower. So we're already, by not taking this into account underestimating flood risk and that underestimate only gets worse in the future. So we're going to try to make sure that we've got something more reliable than just the assumption of a non-changing climate.
[00:12:15] Bridget Scanlon: That sounds fantastic John and I'm delighted that it's so data-driven and then also using the historical records and optimizing the modeling and the observations to have a consistent field. Because I think people, if you can talk to them about what has been happening, and that will amplify in the future, then I think maybe they will have more confidence in it. So I'm looking forward to seeing that information.
Another thing that you mentioned in the legislative presentation was that flash flood warnings in Hunt Texas downstream of Kerr County floods, that there were about 110 of those over the past 23 years. So maybe people get a bit immune to these warnings. And also I think you were emphasizing that warnings need to include some indication of action, should you stay where you are, should you go to higher ground? When should you do it? And in order for people to take heed of these things, the forecasts need to be reliable. And so I'm just wondering, do you think we have the skill to suggest the actions and to improve the warnings?
I know from my own experience after Hurricane Rita, people, a lot of people moved out of Houston and came up to Austin and I was in a wreck that weekend and the person behind me had just come from Houston and they were complaining that their wife made them move and then they ended up being in a wreck.
But also just looking online, they said, a hundred people died and it was a deadly evacuation. So it's a difficult process in a lot of psychology and other things involved. So what are your thoughts about that?
[00:13:52] John Nielsen-Gammon: The whole point of the National Weather Service issuing a warning is because there's something going on that people ought to pay attention to and may need to act on. If there's a tornado coming, find a secure place, take shelter, that sort of thing. But on the other hand, as a society, we push back on rules and being told what to do. We like to make our own judgements. That's certainly true for me. If I see a tornado warning from my area, I'm going to look at the radar. I'm going to look out the window, and I'm going to figure out what I think is happening. Of course, I've got the scientific expertise to be able to make good judgments on that, but it's a natural response from people to not necessarily trust what they hear, especially if it's non-specific sort of information.
And the National Weather Service for flood warning is issuing a warning for substantial part of Kerr County, in this case, a hundred square miles. If anybody who's traveled in the Hill country knows there's lots of what we call low water crossings, where there's the road is designed for the flood waters to go over it rather than under it, rather than building a large bridge.
So if you've got a minor flash flood, the appropriate action is to stay put, not drive, not risk getting washed away at a low water crossing. If you've got though a major flash flood that isn't that. Only one out of the 110 warnings that we've had in the past 20 years in Hunt, then the proper action may be to get to higher ground.
But it depends on where you are again. If you're not at risk from the flood, you should stay put because lightning is a hazard. So that's, that sort of information is not something that a typical weather warning can possibly convey. They can give blanket advice, but the information that's really needed for action is very location specific.
On the one hand that could be handled on a individual personal level by people being very aware of what their flood risk is in a given place and understanding what's possible. But that's a challenging thing to do, to be aware of whatever hazard is and whatever action should be at any given level.
So I think where we'll move to eventually, not in the next couple years, but eventually, in addition to a better mechanism of getting the warning out to people through sirens or whatever, is personalized warnings. Where your cell phone knows how, what, where your location is. That includes your altitude.
So the height above the river is something known by technology and potentially a flood warning can provide the information as to what the potential height of the river is. And your weather app could translate that into a, you need to find a place that's at least 10 feet higher than where you are, for example, personalized weather warnings. Very location specific. We don't have such a system in place, but other than making sure we have adequate cell phone coverage, there aren't actually any technological barriers from making that sort of thing happen. So I'm very optimistic that we'll solve that problem in the next decade or so.
[00:17:11] Bridget Scanlon: Right. And you mentioned people being aware of their risk from floods and stuff like that. When I was reviewing, when the floods happened in the past, a lot of them happened during holidays. And people are visiting like, the Wimberly floods on Memorial Day, the recent July 4th flooding.
But I also heard about Norfolk, Virginia in the Waze app. It provides information on what roads are passable or things like that. And then the drivers then confirm or send feedback back to the system. I don't know much detail about it, but it also includes tidal issues and things like that. But I agree, but then if you lose cell coverage, then you are in deep trouble.
[00:17:53] John Nielsen-Gammon: We could do a lot with local monitoring also, so aren't necessarily relying on an accurate warning. If you are, if you know that the water is within three feet of your foundation, there's a risk. And there, you don't necessarily need a high technology solution for that sort of thing.
There was, at the same hearing in Kerrville, one of the survivors of the flood testified that she was awakened by an alarm going off, and that alarm was their water softener system, which had a, a leak detector. Except in this case it wasn't detecting a leak, it was detecting the fact there was water covering the floor of their house already.
And that probably saved their lives and probably, possibly the lives of some of their neighbors. As long as there's some way to get the information of what's going on, it doesn't need to be from a weather service to emergency manager or an app to the phone. It could be something very simple.
But the problem here was people did not have that information through any mechanism, and they were sleeping until it was too late.
[00:18:54] Bridget Scanlon: And another thing that you brought up in, in the presentation to the legislature was the need for increased monitoring. And the very limited monitoring you described early on about these systems that come together. For Kerr there were like a tail of a tropical storm and then a hurricane from another region.
And you indicated that some of the monitoring at Del Rio, which is 700 miles away, was very helpful. And there's a lot of uncertainty in where the storm is going to happen. And so having more monitoring at the border between, because a lot of systems come from Mexico to the US or, maybe you can elaborate on that, John.
[00:19:29] John Nielsen-Gammon: Yeah it, it's a sort of a weird thing with the atmosphere where you can get by with relatively far apart observations of wind and temperature and pressure because all those factors are tied together, and it's possible to figure out what's going on in between. But water vapor, that the moisture in the air doesn't behave like that.
It can have a lot of fine scale structure. You can see that. You look out the window and clouds in one place and not another. And they're, in this case it was where the moisture was largest, that made, presumably made all the difference. And in the West coast they have a similar sort of problem because they get their rainfall and their flooding from what are called atmospheric rivers, large plumes of moisture that are fairly narrow. They're coming in from the Pacific. And you don't have any weather balloons upstream over the Pacific. Back in the sixties and seventies, there actually used to be one or two ships stationed out in the middle of the Pacific Ocean for the purpose of taking observations and launching weather balloons.
And that was, that was very expensive. They stopped doing that when they got satellites, but you lose the ground truth information, the direct measurements of what's going on. The view from satellites is fairly blurry.
So they've tried to solve that problem by not only sending out aircraft to make monitors, but also basically putting up a sort of a fence line of observing systems, measuring the wind, measuring the humidity and so forth. Since they can't rely on upstream observations.
In Texas, we have sort of the same problem. Half of our incoming air during extreme rainfall event coming over the Gulf of Mexico, where we don't have any observations of what's going on above the sea surface. And then half it's coming from perhaps northern Mexico where there are observations, and we help subsidize them, but they're not as reliable. There were a couple observations missing in, in key areas. So that means, between Brownsville and El Paso, there's only one reliable upper air site, and that's the one at Del Rio, like you mentioned. So potentially we can do much better by having just three or four additional observing sites along the Rio Grande. Probably want to add one or two along the Gulf Coast as well.
And before we just go in and commit to those observations, it's possible to actually run simulations of past weather events and see what sort of impact those types of observations would have and essentially figure out what do we get the best bang for our buck? Where can we get observations that would actually provide key information?
And how much information do we need? Maybe I'm wrong and we actually wouldn't have done better because it is just too chaotic a system, but we can actually test that out and make the investments to to save lives and provide more advanced warnings of the potential for very heavy rainfall in areas smaller than just West and Central Texas.
[00:22:25] Bridget Scanlon: Oh, I think that's that would be fantastic. And I agree, to use modeling then to try to target optimal locations and see how much you would have benefited from that additional monitoring. That would be very good analysis to do. And so jumping to the other extreme, droughts is the other thing that we often experience in Texas.
And I guess the 2011 drought was the worst one year drought in many regions of Texas, when most of the state was in exceptional drought. Maybe you can describe a little bit John, the development of that drought. And I know Nelun Fernando and her colleagues did a nice paper on what caused the intensification of the drought and then the following winter in 2011, the demise of the drought.
I also was looking at, it is very obvious the impact of that drought. And until 2014, you could see it in the reservoir data and Texas Water Board Water Data for Texas, which I think is best thing since sliced bread. You can see the historical reservoir storage, so it's very interesting. So maybe you can describe a little bit about that.
[00:23:29] John Nielsen-Gammon: Yeah, I love that website too. "waterdatafortexas.org". I use it all the time. You can look at current reservoir levels, you can look at how they've changed over the past year. You can look at historical patterns and so forth. And if you do that, you look at a reservoir in West Texas that basically the waters are dropping most of the time. But every once in a while, there's this big upward spike when there's a heavy rainfall event or a flood that cause the levels to rise or maybe even fills up the reservoir. In East Texas, you get a fairly different picture. You see the reservoir levels going up and down, always staying pretty high.
Frequently they're stuck at the storage capacity. So in, in a sense, what that amounts to is normal rainfall in West Texas corresponds to a decline of water supplies, whereas normal rainfall in East Texas is perfectly adequate to maintain supplies under the current amount of water use that we have.
So, per drought, obviously drought starts when it stops raining. People talk about drought early warnings, but that's pretty challenging to actually pull off because you don't know that it's going to stop raining for two months until it has stopped raining for a month or two. Maybe after a month and a half you can say uh-oh, we're reaching the point where we might start seeing impacts agriculturally and so forth.
So at least the agricultural warning levels might be useful in forecasting, but you really can't do much better than that. Our main tool for long range forecasting has historically been looking at things like El Niño and La Niña. Those are temperature patterns and weather patterns of the Tropical Pacific that affect our weather because they're affecting mainly the strength and location of the jet stream.
With an El Niño, the jet streams probably south, we tend to get lots of rain during the wintertime. La Niña, farther north. So we tend to miss out. So in 2010, 2011, that was a La Niña year. It was a stronger La Niña than anything we've seen since then. And that tends to lead to drier than normal conditions, and it definitely did, in this case.
If you knew it was going to have that big of La Niña, you probably would've expected rainfall totals to be maybe 20 or 25% less than what they normally would be. It turned out rainfall totals were 50 or 60% less than what they normally would be. So that, that low rainfall amount probably couldn't have happened without a La Niña, but it also wasn't guaranteed with a La Niña.
It was unusual, even in the context of a La Niña event. Then in the spring and summer, you you get the sort of feedback mechanism in place where if there's not much moisture available in the ground there's not much moisture that can evaporate into the air and you can't really feed thunderstorms.
And if the drought extends into Northern Mexico, then the air will loft, is going to be dry also because the air's coming off the plateaus and essentially preventing thunderstorms from forming, putting a lid on convection. That's a fancy scientific term that we actually use, lid. It means what it says, it means basically the air that wants to rise, can't because it hits warmer air aloft and ends up sinking back down.
So you can get that self perpetuating process. And again, that contributed to it, but also the bad luck of the weather. Circumstances contributed also. So for the first few months the impacts are mainly agricultural. The ground, the land surface dries out fairly quickly. Fortunately, it started in, in the cool season, so it didn't dry out much, but by the time it got into the spring, things were pretty dry.
And normally you'd be getting lots of rain then, but we didn't get that. As the drought persists, then you're just having the reservoirs being drawn down and then you start getting what we call a hydrologic drought where water supplies are affected. For surface water, it takes, generally takes a year or two because usually we've got a several years water supply.
In any given reservoir for water supply droughts, that real ground water it's actually, as you know, a bit different. It tends to be a capacity driven drought, where people need to use more water to irrigate the landscape, irrigate their gardens, and the pumps just can't keep up.
And if some pump goes out of commission, suddenly you're left with a water shortage. It was a very fast developing drought. Then the water shortages kicked in and in much of Texas, the reservoirs didn't fill up again until 2015, even though we had, fairly normal rainfall in 2012, 2013, 2014.
2015 was the exceptional event that caused flooding and caused the reservoirs to fill up. And Wimberley Flood was the most damaging flood, but there was massive amounts of rain in many other parts of the state. In fact, month of May 2015 as a whole, was by far the wettest month on record overall for the state of Texas.
And that basically ended the drought. We had a hot summer, a dry summer, and then another wet fall. And it's the story of the weather in Texas. You've got droughts, then you got floods, then you got droughts, then you got floods. And sometimes you have 'em both at the same time.
[00:28:49] Bridget Scanlon: Yeah, I think that is fantastic. Another website that I also really like to look at is the US Drought Monitor, and we have the Texas Drought, also the Water Development Board host the website on the Texas drought. And what's really nice about it is that you can see the long-term context.
How bad is this drought relative to previous droughts? And so if you look at the data for Texas as a whole, you can see 2011 was off the scale. But then the drought persisted until 2015. My colleague Mike Dettinger has written a paper about the atmospheric rivers in the west and talking about drought busting floods.
And you said you'd normal rain maybe 12, 13, 14, but you didn't fill up the reservoirs because I guess the soil moisture, the soils were so dry that they just absorbed any water. So the runoff was limited. And then you need an extreme event almost to overcome that long-term deficit.
[00:29:42] John Nielsen-Gammon: Yeah we seem to see the same thing. Just recently. The Highland Lakes: lake Buchanan, lake Travis, along the Texas version of the Colorado River, they were very low. They were somewhere between 25 and 50% of capacity, even though rainfall in the, those river basins have been close to normal, but it seemed like it was basically a lot of light to moderate rainfall events that soaked into the ground rather than any heavy events that, that run out, ran off and filled the reservoir. With heavy rainfall in July, that, that problem was solved basically, and some stuff in June. So they got enough rain in the flooding, and unfortunately, you can't rely on a flood because the timing is so erratic.
I think much of West Texas is playing water roulette if they're relying on surface water supplies because you've got this, the weather has this gun that is firing bullets. And sometimes those bullets hold water, but most of the times they don't. And you never know when you're going to get the right bullet is what it amounts to.
We participate in the US drought monitor development, by the way. We have weekly conference calls where we coordinate input to the drought monitor and give our representation as to what we think conditions are like based upon numbers we can point to. Like rainfall's been low for the past two months, as well as our knowledge or information from on the ground of what sort of impacts are happening difficulty with crops growing, or the water levels or stream flow or that sort of thing. And as part of the Southern Regional Climate Center, we're expanding our tools for monitoring that. We're pretty soon we're going to go live with a tool that separates out the rainfall into, the amount that's contributing to runoff and the amount that's soaking into the ground, so we can differentiate between the rainfall that's great for agriculture and the rainfall that's great for water supply. And again, look at those things in a historical context. So we're looking forward to rolling that out in the next month or so.
[00:31:43] Bridget Scanlon: That's fantastic. And is that based on intensity then and the antecedent conditions? I mean, If it's been dry for a while, then some of those soils might be like cement, so it might run off and, I bet it's pretty tricky.
[00:31:56] John Nielsen-Gammon: Yeah, I mean we're doing a crude version where we're just sort, assuming a sort of a standard, typical type of soil. But we'll have the hooks in place to really tailor it to individual locations as we move forward. But yeah, it's based upon something that's developed by US Department of Agriculture for the sake of understanding how much moisture is available for crops. And so the idea is a little bit of rain doesn't help anybody. It gets the top of the soil, gets the leaves, and that evaporates. Then you get some rain and depending on how, what the soil type is, most of that's going to soak into the ground.
But then the more rain you get, the more it's going to run off. And if you've had rain within the past few days, then the greater fraction's going to run off and so forth. So it's not running a full blown land surface model. People do that, and we have that information that's available from other sources as well.
But this is a way of just taking the rainfall and at least getting an initial estimate of what that means for runoff and for soil availability. And when we were testing it out last year with Hurricane Helene, we could really see the difference because they'd had drought in the southeast United States and then Helene came on and you look at the total numbers and gee, rainfall's back to normal.
But the trouble is with Helene, it all fell in less than a week and most of it ran off and the deep soil didn't really get replenished. And so by separating out these factors, we can see those sorts of things that just the raw numbers don't reveal.
[00:33:24] Bridget Scanlon: That's really wonderful. And looking at the US drought monitor 2011 through 2014, and then 2015, it really ends abruptly in May, 2015. When you look at the drought records, and I often see that also in, in the west coast, they'll have a severe drought for three or four years and then it'll end overnight because they'll get a family of atmospheric rivers.
Dumping rain on them. So, that's very interesting. And then after chatting with you the other day, John, and you mentioned that the Highland Lakes had filled up. Then I went back to look at that and saw, they're up 80, 90% full after the rains in July. So that was very interesting.
I saw another presentation that you gave online on drought in 2023. And, you described the drought being initiated by a heat dome and also described the implications for future droughts and you mentioned this Palmer Modified Drought Index. So maybe you can tell us a little bit about that.
[00:34:22] John Nielsen-Gammon: Sure. It seemed like 2023 was going to be a nice wet year. At least it was, it was wet through May and early June, but it seems to have happened and we're starting to see research getting published that, that's looked at. This is, we had this weather pattern set up and the second week of June where, storm systems were carried away from the state and we had a lot of air sinking and warming, and we basically had a heat wave. If you look at the data, and maybe even remember, it was only a couple years ago, it started off very humid. We were seeing very high temperatures, and high humidity. So it was quite uncomfortable, quite dangerous.
As it went last or the week or even two weeks, the humidity started declining. So it was still hot, but it wasn't as humid. And that was great for human comfort, but unfortunately it's a sign that the system, the land surface vegetation is running out of water. 'cause it can't, the plants are not able to transpire, to, utilize water as, as much as they could before because they're running out of it.
And, as a consequence of that, there's this feedback mechanism can take place where you don't have water available in the atmosphere to produce rain. But another consequence is the energy coming from the sun. If there's plenty of water it'll mostly evaporate that water and keep temperatures relatively mild.
But if there's very little water available, the only thing that energy can do is raise temperatures. So when we have a drought, during the summertime, we almost always have unusually high temperatures at the same time and vice versa. This year most of the state had plenty of rain during the summer, and we also had relatively mild temperatures. In College Station where I live, we actually managed not to have any days over a hundred degrees, which was nice.
And 2023 was this sort of unusual setup where a temporary weather event in June had really a lasting effect on our drought situation, until temperatures recovered in the fall and rainfall recovered. But that's part of why it's really hard to forecast out because until that weather event looked like it was developing and maybe a week in advance, it was looking great.
We, we were going to have a great summer.
[00:36:53] Bridget Scanlon: And so this heat dome then amplifies drought? I haven't heard that expression used very much, a heat dome. And then the Palmer Modified Drought Index. How did you modify the Palmer Drought Index to get a better data.
[00:37:08] John Nielsen-Gammon: We didn't modify it ourselves. It's actually an official index. So the primary drought index was originally created, essentially was, it was developed in using data in Kansas. It was designed to be fairly universal but it doesn't. It didn't necessarily take into account how drought develops at different speeds in different locations.
It didn't necessarily take into account how the impact of temperature on drought, on evaporation of different, in different locations. And so various, different versions of the Palmer Index have come into existence in, in recent years and recent decades to try to deal with the shortcomings of it.
And one of them is called the Palmer Modified Drought Index. And there, there's a hydrologic drought index. There's a Z index, which sounds kinda like a Marvel Cinematic Universe index for superheroes. But in any event there are multiple ways of dealing, measuring drought, and you need multiple ways because drought impacts different things differently. If you have dry conditions during a certain two week period during the development of corn, it'll impact the yield whereas other drought during other periods don't. So it's really very situation specific, location specific, and it is definitely not a, there's no perfect index and there never will be.
[00:38:38] Bridget Scanlon: Another source of data that I really like is these reports that you guys put out at Texas A&M, I think it's the Texas Water Resources Institute, and they're titled TX H2O Reports. And they had one on drought and they had a timeline of all the droughts since the early 1900's.
I really like those reports. So mention different droughts and I was looking at each of the droughts that they described and then looking at the annual precip data for the state. I know this is a very crude analysis, but I mean, there was like 1915 to 1918 drought and then 1919 was a very wet and USGS water watch data had very high runoff after that.
And the same for other droughts, like 1938 to 1940. Then 1941 was the wettest year on record in, Texas. And runoff was also extremely high. The 50's drought that most people are familiar with, '57 was very wet. Is there any way to predict when a drought might end and and or when we might go from a, a dry to a wet period or a wet to a dry period?
Or what could we could do to try to improve that skill?
[00:39:47] John Nielsen-Gammon: We can hope for the end of a drought and if we have something like an El Nino, which essentially to weather the normal conditions, then we can maybe even bet on the end of a drought. But because of the importance of flooding and ending droughts, especially water supply droughts it's a big challenge.
We didn't know where the drought would be improved in July 2025, even two days in advance. And after July, 2025, the drought was in, it was almost solved, shall we say along the Colorado River. It was improved a bit near San Antonio.
You go farther west some of the rivers there that provide water for the Winter Garden area and for Corpus Christi they hardly got an uptick at all. Same challenge in forecasting flash flood applies to forecasting ends of drought because they're often the same thing. I am, I'm going to remain pessimistic about forecasting end of drought more than a week in advance.
[00:40:55] Bridget Scanlon: Well, we can keep trying, I guess, and we can look at past records or whatever. But another report that you did recently, the Assessment of Historic and Future Trends of Extreme Weather for the Texas 2036 group I really enjoyed it. Comprehensive, informative and very balanced.
And maybe you can describe a little bit about what you learned from doing that report and why 2036, and a couple of things like that.
[00:41:22] John Nielsen-Gammon: Yeah. Thank you. Yeah, it's sponsored by the organization, Texas 2036, which is interested essentially in long-term planning and solving long-term issues for the state of Texas and its prosperity. And so part of that is understanding whether we're going to have to deal with different types of weather hazards or levels of weather risk in the future.
We do in the past for water supply and all sorts of other reasons. So they contacted me several years back to, to essentially do an analysis of historic trends. And we've updated a couple times since then. They wanted to focus on 2036 because it's in their name. And it's in their name because it's the bicentennial of Texas independence.
But it also is a convenient, it's only 10 to 15 years in the future depending on which report we're talking about. And that's about the limit you can safely extrapolate trends on their own without more and more on climate models for projections. And they had a suspicion, which I agreed with, that a lot of Texans are suspicious of climate models and would find observations be the most reliable.
So we wanted to focus on that. Of course, I, myself as a scientist, I'm able to look at climate models and see what they're saying and understand the reasons for them. Under the hood, we double checked and made sure that the trends that we're seeing were consistent with what projections would show.
But we looked at basic temperature and rainfall, but then focus on extreme weather, extreme temperatures, extreme precipitation, tornadoes, hurricanes, floods, et cetera. And, we can say, you often hear in the news the blanket statement extremes are getting worse. And people tend to think that means all types of extremes are getting worse, but that's obviously not the case. Cold extremes are getting better because temperatures are rising so they don't get as cold as they used to. There are other types of extremes. Tornadoes is my go-to example where we really don't have much of a clue whether they have changed in the past because our means of estimating the number of tornadoes and their severity has changed so much over time.
30 years ago people tended to flee from tornadoes rather than chasing after them. But nowadays, in the era of social media and so forth, people make a living chasing tornadoes. So it's pretty rare that a tornado gets missed, it is not observed. So the number of this has increased tremendously in the historical record, but, the actual number tornadoes in the atmosphere probably hasn't, or we don't really know. Conversely, if you look at tornado severity, the raw numbers say that severity of tornadoes has declined in over the past few decades, but the way we estimate severity is by the amount of damage they produce.
And one thing we're pretty good at as a society is learning to correct our mistakes. And if a tornado causes a certain type of damage, let's reinforce the roof. Let's make our building stronger and let's not be as vulnerable to tornadoes. So we expect the amount of damage from tornadoes to go down, even if the number of potentially damaging tornadoes stays the same.
So in our report we look at all sorts of types of severe weather, and some of them we can say, yeah, this is happening. And some we can say this might be happening, and others we say we've got a few hints as to what might happen. But that's about it.
[00:45:00] Bridget Scanlon: I really enjoyed the report. Temperatures are rising and you explained that in a very simple way. That people could, understand is the number of a hundred degree days doubling between now and 2036, and with a small increase in temperature, so I think people are aware of that.
And so that's a nice way of describing the temperature changes.
[00:45:21] John Nielsen-Gammon: 100 degree days are not just, that's not just large scale climate change. We've got the heat on and the effect that contributes to that where people live or more people you've got less natural vegetation. You've got people generating heat through air conditioning. Air conditioning cools inside, warms up the outside.
So on the one hand, gee, we ought to exclude the heat out effect, just look at climate change. But what we're looking at is what people actually experience. And that is climate change plus the heat out of effect. So both of those are contributing to a number of hundred degree days doubling, tripling, even quadrupling over several decades at the current rate of increase.
[00:46:04] Bridget Scanlon: And what was nice about the report also is the estimation the frequency of extreme rain. You say 10 to 15% increase relative to the last couple of decades. And so it's nice to put things in context. And so the report is excellent and I will include a link to the report on the website and to other papers that you have done, John, to provide information related to these.
But also, people normally say everything is getting worse, and so I really like your balanced communication style. Extreme weather in the winter will probably be less because of the temperature increases. It seemed like we never hear about anything improving very much. So it's nice to have the balance. And the drought, and soil moisture deficits, linked to the higher temperatures. And you nicely explained, even if precipitation is not changing so much, the soil moisture is lower and there's larger soil moisture deficits. So droughts were maybe amplified.
[00:47:04] John Nielsen-Gammon: Yeah, the net total amount of rain is a sort of a big wild card because climate models don't at all agree on the trends in Texas, and it's not really tightly constrained as simple changes in the climate system. But the other things going on are pretty more obvious, like rising temperatures means that you have greater evaporation rates, and drought can develop faster and impacts, especially for soil moisture can develop sooner and persist longer. So even if rainfall stays the same we expect to see worse droughts and more frequent droughts. The impacts of that may vary, also. The climate change carbon dioxide increases can make plants more, more resilient to more efficient in using water.
But they might just, end up growing more and then the bigger plants are just as vulnerable in the future as the smaller plants for today. So who knows what's really going to happen, but at least we can look at some of the factors involved.
[00:48:03] Bridget Scanlon: Thank you so much John. I know you have about three different jobs that you've managed and I really appreciate your taking the time to talk on the podcast today. John Nielsen Gammon is the state climatologist of Texas and Regions Professor of Atmospheric Sciences at Texas A&M, and Director of the Southern Regional Climate Center.
And I think you will understand why he gets the award for being an excellent communicator. So thanks a lot, John.
[00:48:29] John Nielsen-Gammon: Thanks very much. My pleasure.
