I started the day in Liberal, KS. My initial target was between Coldwater and Medicine Lodge. This was well east of what was then the forecasted dryline position but the HRRR and some other convection-allowing models (CAMs) showed storms firing there, farther east rather than near it, and the parameters were forecast to be good, with good shear and greater instability shifting into that area as the system moved northeast. Below is a surface image with my hand analysis of frontal positions at 9:30 a.m.:
I tried to draw the warm front red, but computer did not cooperate, insisting on making it the dryline color. But you get the idea. The warm front was expected to move NE and the dryline to the east. Ended up neither did so nearly as much as forecast. In the end, I would have been better off had I just assumed they would stay where they were at the time of this diagram, rather than moving as forecast by models and most forecasts. That would have actually given me a good target area relative to where the tornadoes occurred. Instead of moving as the models and forecasts indicated, things kind of stayed more like it was in that diagram - the features moved a little, but not much. Instead clouds, cool air, and drizzle spread over SC Kansas - largely an effect of storms to east, but this push of cold, cloudy conditions slowed down progression of the warm front and probably the dryline, too. At one point I looked at a satellite loop and saw that the western edge of the cloud shield I was under was actually moving west. That is probably shen I should have moved to the south or southwest. The models were totally off on storm placement, too. The HRRR was way north and east all morning from where the storms actually ended up developing, and had more storms than there turned out to be. And it had the storms not on the dryline, but well east of it in Kansas and far northern Oklahoma - also wrong as it turned out, since the few daytime storms that ended up developing were close to the dryline, actually beginning in the eastern TX Panhandle.
One after-the-fact thought on the map above - looking at this now, in hindsight, I can see whay there were tornadoes near Denver, although I was surprised when I heard that and it was not on most people's radar. But that low is right where it should be to get tornadoes in that area, though, again, the fact it did not move too much probably helped with those tornadoes. There were some expectations for tornadoes farther east in Colorado, but not so much near Denver.
I went east from Coldwater to Medicine Lodge around 1 - a mistake - due to expected movement of the dryline, which did not happen, along with the incorrect model depiction of storms developing farther east. Then the HRRR (which was wrong all day) showed storms developing to my SW just over the KS line in OK, and moving east in OK, as shown in this image showing predicted reflectivity at 21z (4 p.m.):
This was still too far north and somewhat too far east compared to what verified, but at least it did show a shift toward the south and west from earlier runs. So I blasted south, but too late and not urgently enough. Instead of OK, storms fired in the northeast TX Panhandle, which still remained east of the dryline. When I saw a good-looking supercell there on radar, I decided to go for it, but too late and was too indecisive about my route and took too long. This was the storm that produced tornadoes - a few chasers smarter than me got incrtedible video. The storm also produced up to 5-inch hail. But even compared to the HRRR forecast shown above, this storm was more isolated (one intense supercell with a few left splits, but never a chain of supercells like the HRRR prediction), and both farther west and around an hour and a half later. In hindsight, I wish I had followed my gut instead of the models, and as usual, I should have been more decisive. Easy to say in hindsight, and always easier to say afterwards than to do in real time.
Finally caught up to it west of Vici, but by then the Arnett tornadoes were over, the storm was weakening, and I could see no structure due to haze, which remained quite thick. Not sure of the time by then, but would guess it might have been around 7. The view was so bad I did not even bother trying to take pictures. After that, I chased it north on back roads east of Vici and Sharon and finally goot a good view of the updraft. The main cell passed ahead of me just NE of Sharon, moving ENE and weakening rapidly. Another more impressive piece of the storm was to the west - some kind of split but maybe not a true left split as it did not look anti-cyclonic. But it did have nice barberpole LP supercell structure and I got good views of it from a little south of Freedom a little after 8 p.m., just before it produced golfball to 2-inch hail west and NW of there:
This was a pretty storm, but hard to see through the haze, at least from where I was viewing it, along with loads of other chasers including multiple tour and college groups. By this time, there were tons of chasers, as more and more who had come down from Kansas joined in the hunt. At least I got to see a nice LP storm, and on many chases I would have been very happy to get a storm like this, although maybe with less haze. But it did not make up for missing the earlier tornadoes - some of the chaser video of those was amazing. Big picture: some chasers got great ideo of them; many others like me were too late coming down from Kansas. Most of us ended up in the massive convergence south of Freedom, getting the consolation prize of the LP supercell/big-time haier. At least everyone was pretty careful, not doing stupid stuff like parking on the right-of-way or setting up tripods on the road. And it was not a huge backup; traffic moved along as well as could be expected.
Ended up in crowded Alva, OK for the night. Fast food joints and hotels/motels got slammed like they probably haven't in a long time. I got the last room where I ended up staying, then a 20-minute wait at McDonald's for a late dinner around 9:30 p.m.
Edit: I worked a little on video I got of the LP supercell near Freedom. Here is a zoomed-in video of the impressive cloud motion in the upodraft (2x timealpse, with music):
Here is a regular-speed showing the whole storm Both storm-scale rotation and upward cloud motion are evident. Regular speed video with music:
Chase distance for this day: 412 miles.
Total mileage for this chase trip (2 travel days and 2 chase days): 1625 miles.