It seems like every chase days starts the same way, with a benign sky, some small fair weather cumulus and high level cirrus. This was actually the first official chase day of my 2010 plains trip, I was supposed to be down two days earlier but I ran into car trouble. The GM Cobalt I had rented jammed with the key in the ignition in the start position locking the car.

The SPC outlook the morning of the 18th showing a moderate risk for severe weather across the Texas Panhandle.

The synoptic setup which gave rise to the Dumas supercell was about as classic of a dryline setup as you get. A strong flow straight out of the Gulf of Mexico coupled with local moisture created a moisture axis about 50km ahead of the surface dryline. The surface environment was strongly dominated by southeasterly winds, a key component which is often necessary in keep storm inflow warm and pushing outflow away from the main updraft. The entire surface environment was perfect, and aloft everything would be strongly sheered by a moderate 50 knot jet-streak coming in from the southwest and exiting to the northeast. This meant any surface convection would be vented to the northeast and away from the general easterly direction of the storms motion.

This was just an hour before the storm formed, I was at a Love's just on the outside of Amarillo. The question now was, would the dominant storm of the day be north of Amarillo or just south? I was watching two blips on radar that were riding the dryline. The big explosive energy was yet to come but the blips were moving into increasingly better surface moisture.

The moment of truth had come, the blip just to the south of Amarillo faded and the blip to the north suddenly broke 30,000ft and was rising quickly. There were basically no radar returns above 15dBz at the time, so everything was still dry and low precipitation in nature. As I approached the storm which was slowly heading towards Dumas the towers and structure began to become visible.

I stopped just outside of the town of Dumas TX and waited for the storm to come to me. I did not want to rush right over and immediately have to "chase the storm". It was only moving at 30 km/h and at this time it was still around 20-25 km away so I knew I had a good 40 minutes to just relax and watch it. I had the HAM radio scanning the skywarn bands as well as the local simplex channels most of the chasers use. Beyond a few brief squiggles of spin here and there there was no tornado reports coming in.

I took the time to setup my video camera and roll a wonderful time-lapse. It was amazing to watch the storms inflow region change and slowly develop as the storm began to take on more of a classic supercell appearance on radar.

This funky looking cloud stretching from the left to right and scraping the ground is known as an inflow tail. Storms will attempt to eat whatever air they can regardless of where it is coming from. In this case an inflow tail is when you have a band of moisture visible as a storm pull in air that is outflow cooled. Essentially the storm is eating a bit of it's own outflow. The same thing happens anytime a wall cloud develops but is often far less pronounced. If there was to be a tornado it would be somewhere immediately in the canter of the image. In fact I was so absorbed with my photos I did not even see the growing chaser convergence around me. When I first pulled over there was no one, I was alone. Within about 30 minutes there were 30 or 40 different chasers around me.

As the storm slowly approached it was cycling through its various phases of development with inflow features developing and then dissipating. The rotation was also becoming much more concentrated and the NWS gradually issued and expanded the tornado warnings.

The Dumas storm as it approached began to reveal beautiful structure and the different colours associated with the hail and rain bands became more visible. It's hard to explain but down in the great plains everything is just so rich, from the detail and colour of the grasslands to the stormy skies, it's all majestic.