Dwight Silverman

This installment of the Space City Q&A is particularly tasty: Y’all came up with some very good questions! Keep them coming: Hit the Contact link at the top of the blog, or leave a query in the Discourse comments below this post.

---

Q. I have a question as to how weather radars generate their one-hour plus future radar loop. Some recent rain showers have been from storms that kind of pop up on the radar, meander around for a bit and then dissipate. They usually don’t have an easily defined direction to them like they might have if they were spawned from a frontal boundary moving in. However, on most radar loops that I watched, as the radar passes from the older, real radar data to the projected data, the storms go from wandering with no direction to a well defined direction. In reality, as time moves forward the storms end up meandering and don’t go in the predicted direction. Any thoughts as to why this occurs?

A. I am not an expert in predictive radar apps, but I have done some research in this area in the past. So I’m going to dive in a little bit here because I think it’s a good question.

First of all, as you’ve suggested, these predictive apps work best when there are clearly defined storm motions, such as along a frontal boundary. This is because the “nowcast” for 0 to 60 minutes is based on current radar readings and inferred motion vectors based on storm intensity, speed, and direction. More sophisticated methods might also incorporate flow equations, such as the Lucas–Kanade method. (Dear reader, it has been 30 years since I took an advanced mathematics class, and I’m not going to try and explain how that works. Because I really don’t know.) These forecasts are quite accurate for storms with clearly defined motions, but less so for other storms. I would trust them for about 30 to 60 minutes out, at most.

Then there are medium-term forecasts that run from 1 hour out to several hours. These generally rely on a single high-resolution model, such as the High-Resolution Rapid Refresh (HRRR), or a blend of models with some AI-enhancement thrown in. Again, for something like a line of storms ahead of a frontal passage, these can be pretty accurate. But if storms are just popping up around the city during the afternoon, they’re probably not going to reliably tell you if it is going to rain at your house three hours from now. I mean, sure, they’ll get lucky some time. But they’re also going to miss. I wouldn’t trust any predictive radar app more than a couple of hours out.

–Eric

Q. It’s not new, but it feels more noticeable lately with the new weather reporter at the Houston Chronicle. I usually read him, check your forecast, and glance at the iPhone app. A recent day offered a good example: he predicted highs in the upper 90s, possibly hitting 100. My iPhone said 89. You (correctly) called for mid-to-upper 90s. How can one source call for 100 while another is calling for 89? I guess it feels like the spread is getting wider, not narrower. I’d expect the opposite with improved technology. I assume it comes down to different models, how showers affect afternoon highs, and maybe even some editorial choices (the Chronicle does seem to lean into more dramatic weather headlines recently).

A. I can’t speak for the Chronicle’s forecasts, but I know Justin Ballard tries to do a good job. And if there are sensationalistic headlines, they’re generally coming from online editors pressured to drive up traffic numbers rather than the forecaster. I dealt with this when I worked for the newspaper, which is one reason why I very much prefer working for myself.

Anyway, I suspect it may be a combination of things. When I say the highs in Houston are likely to be the mid- to upper-90s, I’m referring to the central portion of the area, i.e. Harris County. Daytime temperatures in the summer will invariably be cooler in Galveston, near the water, and often warmer further inland, such as in Conroe or Sealy. Perhaps Justin was referring to the forecast at Bush Intercontinental Airport, and I was more considering Hobby. That would explain the variance in our forecast.

As to why the iPhone said 89 degrees, perhaps it was forecasting an afternoon shower to keep things cooler? I don’t really know. But in general, a lot of the variety in temperature forecasts for Houston are due to its large area (more than 90 miles from Galveston to Conroe), varied nature of the terrain, urban heat island effect for the city’s core, and surface temperatures of the Gulf of Mexico. It’s … a lot to balance.

–Eric

Q: As far as climate change goes, rising global temperatures get a lot of attention, I was wondering if global humidity levels are trending upward, too? This year may be more an extreme example, but I was wondering: As the temperature rises, and more water evaporates from the oceans and other bodies of water, giving precipitation events “more to work with.” i.e. Has there been a noticeable shift upward in global average humidity levels?

A. When you think about humidity here in Houston – or “houmidity,” as we sometimes call it – you (should) think about dewpoint. Dewpoint is a measure of moisture in the air; it’s the temperature that the air would need to cool to for it to be saturated. When dewpoints are high, it feels very humid, hence why Houston’s summers are so generally awful; we have dewpoints often in the mid-70s to even upper-70s at times. Periodically, the immediate Gulf Coast can see dewpoints of 80°, and there have been instances of some places in the Middle East (Iran in particular) seeing dewpoints over 90°!

Why does it feel so humid? Warmer air holds more moisture. You could have a relative humidity of 100% with a temperature of 50° and a dewpoint of 50°. That feels damp or clammy, but it isn’t oppressive. But if you have a dewpoint of 75° and an air temperature of 75°, that feels disgusting. The relative humidity is 100% in that case too. But because the air temperature is warmer, it’s holding more moisture.

Thus, as climate change continues, we will see global humidity levels increasing too. A warmer planet has more capacity to hold moisture, so generally the humidity and dewpoints will be higher. You can read more from the UK Met Office here. Thinking about this locally, 2025 sure seemed like a fairly mild summer by almost anyone’s standards in Houston. But it was the 6th warmest on record because our nighttime lows were second highest on record. Put simply, the daytimes weren’t so bad, but the nighttimes were persistently very warm, even by Houston standards. Some of that is certainly urban heat island affect (more concrete = warmer nights). But it doesn’t explain why Bush, College Station, Galveston, and Hobby have all had top 20 or 30 hot summers every summer since 2016 (with the exception of 2017 in Galveston). That is likely because the atmosphere is warming, the Gulf is warming, and it’s able to hold more moisture and keep nighttimes much warmer than usual.

–Matt

Q: Are the CenterPoint Energy weather stations available to the public? If so, how do I locate and see the one near me?

A. Yes they are!

All the data from the sensors is being pushed out publicly through a company called “Synoptic,” which acts as a data source for various sources of weather observation data. The data can be accessed through a NOAA website (that’s a fantastic resource beyond just CNP weather stations). Within the Weather & Hazards data viewer, you’ll be presented with a buffet of sites you can choose from. These are not just CenterPoint weather sensors, but they include everything from Harris County Flood Control to Union Pacific Railroad to major airport sensors like at Bush and Hobby. You’ll have to click around a bit to find a CenterPoint one, but it’ll be a number 1-100, a location name, and then a suffix of CE. You can toggle the variable you want to see, including wind gusts, dewpoint, or rainfall, among others. There’s even a historical data mode on this site if you want to see observations from a particular date and time. It’s just a great site to bookmark!

There are also links to 3 day and 7 day histories as well. Those can be bookmarked for real time access updated every 10 minutes (such as this one for Sugar Land), and I would encourage you to bookmark the ones from any source closest to your house!

–Matt

Q: I live in Tomball off Texas 249/99 and we have had some intense storms Monday evening and Tuesday morning. It got so bad Monday evening that I got my family in the bathroom as winds had to be 40-50mph and were getting quarter size hail. The same thing happened this morning to where the wind was so intense that we had trees down in our neighborhood. I looked on my radar app and at the time, there was not even a Special weather statement. Later on they did issue it after it had passed through. I have seen storms that were severe warned that were much less intense. How a storm can be that intense and the NWS didn’t have any warning on it?

A: I’ll give you a good example of this. I am assuming that your experience was during the storms we had near the end of August. One of them produced a 66 mph wind gust at Brazos Bend State Park (from a CenterPoint sensor actually!) at 5:10 PM on August 21st.

Looking at radar, I’m not especially close to issuing a warning on that storm based on what we see above. One issue with Houston is that we are such a sprawling city that we actually have some pretty healthy data gaps when storms push through the area. Radar is critical, but it is also imperfect sometimes. Even the higher resolution and slightly closer Hobby Terminal Dopper Weather Radar showed winds of only about 35 mph in the above example. No indication whatsoever that this storm is producing 60 to 70 mph winds, but it is.

So there are instances where storms can (literally) fly under the radar. The radar beam up your way is shooting up to around 3,500-4,000 feet above the surface. So there is a chance that in those cases it was missing what was occurring closer to the ground.

It’s also not granular in time, either. Periodically, something will happen with a storm where it does its dirty work over the course of a minute or two, and it could happen in between the 2 to 4 minute radar scans we often get. Also, storms here tend are often what we call “pulse” storms: They ramp up quickly and ramp down just as quickly. You can literally go from a shower to a 50 mph hail-producing storm in the course of minutes. If I am correctly recalling the day you’re discussing, those storms were in fact remarkably quick to rev up. NWS did issue warnings, but there was a bit of playing from behind going on because of the nature of those types of storms.

But the reality is that until we get to more substantial radar coverage at more granularity than we’re currently capable of right now, there will occasionally be a miss like this from time to time. It’s the nature of uncertainty. It’s also a good reminder that if you see something, report it! Send it to us, the NWS, or your favorite local meteorologist. We’ll try to make sure it gets to the right people.

–Matt

Once again, you’ve asked and we’re answering. Here’s our latest Q&A. Got questions of your own? Hit the Contact link at the top of the blog, or leave a question here in the comments. We’ll consider it for next time.

---

Q: I’m not complaining at all, but it seems like this summer has been cooler than usual? I know we’ve had a lot of rain and maybe that’s kept things cooled off but is something bigger going on? Is there an overlying thing in the atmosphere that’s made June and July less than hellish?

A: Good question! We’ve actually already answered it because we felt the same thing! So, ok. Let’s look at this again with July now in the record books. Through the end of July it was officially the 7th hottest summer to date on record in Houston dating back to the late 1800s.

These temperatures are taken at Bush Airport, which about 20 percent of our audience thinks is complete blasphemy, another issue we addressed. And while Bush does have problems from a representation of Houston, when comparing to prior history at that specific location, it suffices. But, for those who would rather look at the data elsewhere as comparison, let’s do just that. At Hobby, it was the 15th hottest June & July on record back to 1931. In Galveston it was 4th hottest. College Station? 25th hottest. Corpus Christi was 9th. Beaumont was 44th! So, suffice to say there’s been variability across Southeast Texas. Was this really the 7th hottest start to summer “for Houston?” Officially, yes. In reality, probably not quite.

But it really depends on exactly where you are. And this gets to the reader’s mention of “a lot of rain.” Bush saw 9.46″ of rain in June and July, much less than last year of course, thanks especially to Hurricane Beryl. But if you took Beryl out of the equation it’s been a couple summers since we’ve had this routine amount of showers and storms. Hobby has had about 4 inches more rain this summer. Galveston has had far less, with about 4.5 inches officially. So it seems like the rankings of “hottest” are better correlated to rainfall than anything else. If you’ve seen more persistent daily storms in the Houston area, it’s been a good summer. If you haven’t, it hasn’t been horrible like 2022 and 2023, but it’s been a little hotter.

One thing I will note, that we’re basing this on average temperatures. When you look at nighttime lows, we’ve had more mornings at or above 78° this summer than in any prior summer in Houston. This persistence of warm nighttime lows is making a large impact on our overall average. This is for a number of reasons: A warming Gulf, urban sprawl and expanding heat island, and climate change. It’s just easier to have hot summers in Houston now than it used to be, even if the daytimes aren’t as bad as we’ve dealt with before.

– Matt

Q. Seems like we’ve had several low pressure systems near us (near shore) in the Gulf that IMO has really helped us with rainfall since June. Any chance these little low systems have pulled energy out of the Gulf that would keep something bigger from happening the next month or so?

A. The short answer is no, sorry. For two reasons.

First of all, it takes a stronger storm, preferably a hurricane, to really churn up lots of colder water from below toward the surface. This raising of colder water can produce a lingering effect on sea surface temperatures that reduces the energy available for subsequent storms to intensify. None of the weak systems we’ve seen to date have really been potent enough to have this effect.

Second, the effect is temporary, usually lasting on the order of two to four weeks. So even if the June tropical systems had a significant cooling impact on seas (which they really didn’t), it would be long gone.

Speaking of Gulf sea surface temperatures, they are plenty warm. The vast majority of the Gulf lies at 30 degrees Celsius or higher, rather far above the threshold (roughly 26.5 degrees C) needed to support tropical system intensification. Buckle up!

– Eric

Q: Matt, I know you have written about cuts to NOAA and NWS, and you’ve said the forecasting and warning around the Hill Country tragedy was not affected. But is there anything now you can point to and say, “That happened because of these cuts”?

A: I stand by my initial assessment. I don’t believe the cuts, as implemented thus far played much if any role in this disaster. As we’re learning from the state legislature’s special session, it appears there’s a fundamental problem with emergency response to significant weather in Kerr County in particular. While the NWS San Antonio office was without a warning coordination meteorologist at the time of the flood, the relationship building that occurs to help mitigate the toll from this sort of event occurs over a long period of time. Could it have helped here? Maybe a little? No one knows for sure. Would it have prevented it? No. I feel strongly that this was sort of a terrible situation, wrong place, wrong time, wrong leadership type of situation. Hopefully there are lessons learned from this.

I encourage you to read this by my friend and colleague, Alan Gerard, who really dug into what we know about the flood, the warnings, and the response.

The caution here, of course, is that the cuts that have been proposed to NOAA in the administration’s budget would be catastrophic to NWS and increase the risk profile across the board for events of this and lesser magnitudes. It made absolutely zero sense from a public safety standpoint. Thankfully, both legislative branches are playing the roles of grown-ups in the room and have come back and said those cuts aren’t happening. What this looks like in the end remains to be seen. But to this point, what is being said by Congress is actually net positive for NOAA and NWS and the country at large.

As far as other significant weather events and the current budget cuts, we have not noticed other cases where we can say with authority that the cuts have played a detrimental role in outcomes or forecasts. There was the instance of tornadoes earlier this year in Kentucky immediately after the Jackson, KY, office announced they’d stop staffing 24/7 due to the lack of staff. But the office was open and operational that night. The outcome was suboptimal, and the cuts may have played a role there — but like a lot of things, we can’t exactly prove it outright.

Same goes for the lack of weather balloons. On at least 2 or 3 occasions this year in forecasting for both the Houston area and for my day job in the Midwest, I have noticed a couple “surprise” outcomes that I strongly suspect are tied to the lack of upper air data provided by weather balloons that has been inconsistent from parts of the Rockies and Plains. But again, that’s speculation to some extent. We know for a fact that these are bad decisions, but proving what role they’ve played in any one event is difficult to do.

– Matt

Q. I was looking at my records and thought “hmm, haven’t done my annual contribution to SCW yet, let’s find a link on their website”. But no, couldn’t find one. Are donations only while you’re having your annual thing, or did I miss something?

A. You’re very kind to ask this. We hold an annual fundraiser over the course of three weeks in November. During this time period people can donate money to support the site and buy merchandise for the holiday season. Speaking of which, do you have any ideas for a t-shirt design this year? We want to do something fun for our 10th anniversary.

Why only November? We don’t have a year-round fundraiser because I don’t want the site to be seen as trying to capitalize during inclement weather events. In November, things are usually fairly calm, so there’s no pressure on people to support us.

– Eric

Q. Your app keeps crashing. What’s happening? I thought you fixed it!

A. Sigh. So did we!

When last we left our cute little weather app, we had released a new version that fixed the longstanding issue preventing notifications on Apple devices. As we wrote in June, it took a while to get this done because Hussain Abbasi, our developer, needed to switch to a new coding platform – something he wasn’t planning to do until next year but Microsoft dropped support for it.

The notification problem was indeed solved, but as is often the case when you’re making a change as major as this, other bugs were introduced. After the new version had been out for a while, we began to get reports of crashes when the app launched. Hussain found the issue, having to do with how the locations in the city-picker drawer get sorted. A updated release with a fix for that is imminent – hang in there! (In the interim, deleting and redownloading from the App Store seems to solve the problem.)

Meanwhile, he’s still in the process of getting the Android version working as it should via the new coding platform. Building Android apps can be more difficult because there are so many different hardware types and designs out there. In this case, you can blame folding Android devices for the delay – switching between tablet and phone layouts is creating, as he puts it, “inconsistencies.” (And there’s supposedly a folding iPhone coming next year. Yay?) Once we get an Android update out the door, it should also fix the issue of crashes on some Google Pixel phones.

It’s like playing whack-a-mole. Whack-a-bug!

Remember that you can send a bug report via the app. First make sure you’ve downloaded the latest version. Then tap the three-line icon next to the city name on the home page; tap the gear icon on the top right of the drawer that appears; tap the “Send Feedback” item. It sends a log with details on the crash to our developer, but it would help if you could provide as much detail as you can about what happened. (Details about your device are automatically sent.)

Does the app crash on launch so you can’t get to the Feedback button? Send as detailed a report as you can, including the device model and its operating system version, to [email protected].

And once again, just to be very clear: The spelling of Houmidity in the app is a feature, not a bug.

We appreciate your patience as Hussain works through these issues.

– Dwight

We’re back with a new Q&A! Eric and Matt have had their hands full with their day jobs the past couple of months, and that is likely to continue, so consider these kinds of posts “occasional” rather than monthly going forward.

Got a question of your own for our next round? Leave it in the comments, or drop it into the Contact link on the home page. We may also troll our Facebook page for Qs that we can A.

– Dwight

---

Q. What does the temperature being so hot this early do to the hurricane season this summer?

A. The average temperature in May was 3.2 degrees above normal in Houston, and nearly 2 degrees above normal in June. Part of the reason for this is that sea surface temperatures in the Gulf of Mexico are warmer than normal. (This influences local air temperatures, especially at night).

A warmer Gulf, of course, means that any passing hurricane or tropical storm will have more energy available to support intensification over water. So in that sense, warmer temperatures are more favorable for stronger hurricanes. However, aside from that connection, there is nothing I am aware of that links air temperatures in Texas in early summer to making the region more prone to hurricanes in a given season.

– Eric

Q. I’ve seen your posts about local forecast issues because of weather service cuts, but I haven’t seen much about how it might affect hurricane forecasting in general. Can you address this?

A. Like just about everything, the real answer to this question is somewhat nuanced. Simply put, we rely on a lot of data from NOAA and the National Weather Service to inform our forecasting. So any cuts to their services or data would be problematic. Despite all the noise around budget cuts and such, thus far, there have only really been 2 significant items of note that could negatively impact actual hurricane forecasts right now.

First is the loss of routine weather balloon launches collecting upper air data across parts of the middle of the country. We covered the impacts of this at The Eyewall in detail back in March. In short: If those weather balloon launches remain on hiatus with active storms in the Gulf, that could negatively impact the forecast of the tropical storm or hurricane track enough so that we would want to “pad” our forecasts by a few extra miles on either side of the cone.

Secondly, news that came down late last week is more important and troublesome. We also covered that at The Eyewall. In a nutshell, a Department of Defense satellite is nearing the end of its life. Last week, they announced data would cease on Monday from a particular instrument on that satellite that is critical to hurricane forecasting. Early Monday it was reported that the satellite instrument had been granted a reprieve until the end of July. This Special Sensor Microwave Imager Sounder (SSMIS) helps meteorologists see through the clouds in tropical systems, allowing them to make much better intensity estimates. The loss of that data will lead to somewhat poorer initialization data on storm intensity and location, which could impact every model we use to forecast. While there are some fallback measures in place, it goes without saying that that is a huge loss. Unfortunately, aside from questionable communication of the decision from the government, this seems like just a poorly timed coincidence more than anything.

Neither kneecaps us this hurricane season in terms of forecasting, but it will make us think twice about being highly confident in any given scenario, even within a day or two of landfall.

Related to this, is that the administration’s proposed budget for NOAA was formally given to Congress on Monday, and it basically destroys the entire infrastructure of weather research in America. That’s a much bigger threat than anything done so far, and it would without question set weather forecasting and research in this country back years.

– Matt

Q. Last year I moved into a house with a yard for the first time in my adult life. I lived in a townhouse for almost a decade and one of the things that attracted me to it was not having to take care of a lawn. Unfortunately I’ve gotten to where I was really struggling to live in a two story. I have someone who cuts the grass and amazingly have been able to get away with not watering yet, but I know it’s coming. So any hints? How do I know when it needs it? How long to run the sprinkler? What time of day is best?

A. Well, you’ve picked a good year to have a lawn. March and April were fairly dry months for most of the region, but we had not yet reached the real growing season for lawns yet. By the time warmer temperatures started to arrive in May, we began to see regular rains. The pattern continued into June when we have seen plenty of rain. So far, so good, in terms of not needing a sprinkler this year. But July and August are looming, and I doubt we’ll get through the next two months without needing to water lawns.

As a general rule, I find that if a yard in full sun receives a soaking rainfall, it can go about 7=10 days before needing to be watered. Areas in partial shade can go several days longer. There is no secret about when grass needs to be watered. First it will start to wilt, then the grass will appear to be thinner and weaker as it dries out. The ground will be hard. This is the last step before grass begins to brown and dry out. You need to water the grass before it turns brown, because then it will be too late.

In terms of amount, I find that 20 minutes every three days works pretty well. Early morning is best. Please note I am far from a lawn expert, but I do have a quarter of a century of experience keeping a lawn alive in Houston.

– Eric

Q. Is it true that flat land is more prone to tornados? I know the area around St. Louis has some flat land. And we do get some tornados here. Does Houston’s urban landscape, with lots of tall buildings, protect us at all?

A. There are a number of persistent myths about tornadoes that have withstood the test of time for one reason or another. First, tornadoes seem to be more common over flat land because a number of tornadoes that are visually documented in the most dramatic fashion tend to occur in the Plains, which are generally flat.

Still can't believe this happened. Still can't believe this is a photo I took.
📍Wellfleet, NE 6/16/25 pic.twitter.com/9X7Vu6ylof

— Jim Tang (@wxmann) June 17, 2025

I mean, it’s tough to beat an image like that. But reality is nuanced! Tornadoes have occurred in all 50 states, and what determines where tornadoes most likely occur is less about terrain and more about large scale weather and climate patterns, which is why the Plains and Southeast tend to lead the country in tornado events. But at a local scale, some areas can be somewhat more favorable than others due to topography. For example, we’ve seen a few examples of tornadoes impacting near Pikes Peak in Colorado in recent years, including just last month!

Tornadoes have occurred even higher in elevation in Colorado including a 2012 Mount Evans tornado at nearly 12,000 feet.

Two places near where I’ve lived have some topographical features that can produce enhancement of tornado risk: Parts of southeastern Pennsylvania have mountains to the west and Delaware Bay to the southeast, which almost acts like a much smaller scale version of the Rockies and Gulf respectively in feeding a miniature tornado alley. In Upstate New York, added “twist” to the wind north of Albany because of the intersection of the Mohawk and Hudson Valleys can add a little extra “oomph” to tornadoes there too, an area that is more hilly than flat.

Houston’s urban landscape does not really protect us. While it may act to locally enhance or “shadow” certain thunderstorm impacts, the downtown skyscrapers are not immune to tornadoes. A tornado went through downtown in 1970. Granted it was fairly weak. But Houston’s tornado history is littered with generally lower-end tornadoes in all corners of the city. Other major downtowns that have been hit by tornadoes include Nashville, Fort Worth, Miami, Los Angeles (yes, for real), Salt Lake City, Atlanta, and more.

So downtown skyscrapers offer no real protection. The reasons downtown areas get hit so infrequently overall is simply a matter of luck and math. The footprint of rural land is much greater than that of urban land. So the odds of an urban center being struck by a tornado are automatically much lower than an open field or farmland in the Plains. And it’s why tornadoes that hit cities tend to get covered significantly in the media, simply because more people and structures are impacted.

Bottom line: Tornadoes can occur virtually anywhere and cityscapes and elevation in most cases don’t offer protection. The causes of tornadoes are because of atmospheric ingredients and geography, sometimes buoyed by local effects of topography.

– Matt