Arellano, who was born and raised in McAllen, is the meteorologist-in-charge at the National Weather Service forecast office in New Braunfels. His career, which began in 1976, has taken him all over Texas, as well as to Puerto Rico and Florida.

There’s an old saying here in Texas: “Either you’re in a big drought or you’re in a big flood.” Everyone knows we’ve been in a big drought lately, one of the worst we’ve seen. We’ve had less rain than in any of the one-year droughts in the fifties, which were considered a benchmark. When you have dry conditions and the kind of high temperatures we saw this past year, the drought feeds on itself.

The dry weather doesn’t allow precipitation to form. That’s why all those storm systems dissipated over Texas in the summer, because of the abnormally strong high pressure sitting over the state. In July, for example, Tropical Storm Don fell apart as soon as it made landfall. When La Niña forms, as we are seeing this winter, there’s a tendency for it to form again over the next several years. And in this part of the country, La Niña means drier weather, which means more drought.

You might think the drought doesn’t bring us work, but it does. There’s information to provide, records to track. And then there’s the danger of fire. We monitor the forecasts and issue red flag warnings when conditions are prime for wildfires. We dispatched meteorologists to Bastrop so they could give the morning briefings to the firefighters, telling them the direction the wind was blowing and whether winds would shift later in the day.

Texas has it all: tornadoes, floods, hurricanes, dry spells, you name it. We have the Gulf of Mexico in the east, which brings a lot of moist air onshore, and then we have dry, arid West Texas. When a cold front comes in, that air mixes with the humid, warm air of the Gulf, producing our classic thunderstorms. Then, in the center of the state, there’s the Hill Country, which has extremely rocky soil that can’t quickly absorb rainfall. The creeks rise rapidly, and you get all the flooding.

I was a middle schooler in the Valley when Hurricane Beulah hit, in 1967. My mother made my brother and me sleep in the closet, away from the windows. I remember how the wind howled. Beulah dropped nearly thirty inches of rain. The levees in the Valley started to fail, and about three days later, at 4 a.m., the fire department knocked on our door and told us to evacuate. We grabbed whatever we could, got in our station wagon, and headed to higher ground.

The water rose only to our front door. But the experience marked me so profoundly that when I graduated from Texas A&M University, in December 1975, it was with a degree in meteorology. I applied right away for a job with the National Weather Service, but President Nixon had just imposed a federal hiring freeze. I worked in the hardware department at the Sears, Roebuck in Bryan for a while before I got a call asking if I was interested in an intern position with the National Weather Service in San Juan. I accepted without really even knowing where Puerto Rico was.

Back then, we still used Teletype machines and radars from the fifties. We’d get a picture from a satellite every thirty minutes, and all our charts were hand-drawn. Nowadays, when we send out a tornado warning, you get it on your smartphone in minutes. You even get satellite images! But in San Juan, I’d have to stand outside at the airport, every hour on the hour, look at the sky, and note the clouds, visibility, precipitation, and fog. Today our systems are automated, but in those days we were always running up and down the stairs, trying to get our observations as quickly as possible. We had big paper mats where we plotted the information—cloud symbols, air pressure, wind speed—into spaces the size of a dime.

After Puerto Rico, I moved to Tampa and did fruit-frost forecasts, which put the citrus industry on alert for freezing temperatures. If the temperature dipped below 28 degrees, the growers would turn on heaters. Or some nights, the cold air would sink, in an inversion, and they would hire helicopters to fly over the groves and get the air spinning. But the Weather Service eventually got out of agriculture forecasts, so after a stint in Brownsville—there’s a lot of citrus there too—I became the lead forecaster in San Antonio, where I covered all of South Texas. Then I moved to Corpus Christi, and then to New Braunfels.

When you major in meteorology, you need to know math and physics to study the motion of air. But only after being in a location for several years can you really understand its peculiar phenomena. Like how, for example, the Hill Country escarpment can lift clouds, produce rain, and change the direction of the wind. Or how, with hurricanes, the question is “Where is the center of the storm going to be later in the night?” Slow-moving storms mean that all the rain will fall in one area; in 1998 Tropical Storm Charley dropped a year’s worth in one night over Del Rio. But anything fast-moving, like a flash flood, can cause extreme damage, so you want to provide the best warnings possible. At the office, we have our own generator so we can hunker down. We have big metal shutters, to cover the windows, and satellite phones.

In the past, we’d give a three-to-five-minute lead time warning for tornadoes and flash floods; now that lead time has grown to fifteen minutes or, in some cases, as much as an hour. It used to be, with our old WSR-57 radars, that you could see storms and how heavy the rain was, but you couldn’t dissect them. The radars had vacuum-tube technology and were so old that some of the parts had to be bought from Russia. But then, in the late eighties, came the Doppler. It is so sensitive that it can pick up bats when they come out at night. If you look closely, you’ll see plumes of them streaming out of their caves. Sometimes, television meteorologists who are not familiar with the area can mistake the bats for rainfall. They quickly learn otherwise.