Testing the blog abilities/functionality, but also wanted to highlight an observation I had regarding the GREMLIN data and its handling of light precip echoes and cirrus.
As expected, the stronger echoes in the line of storms to the west of Minneapolis are handled very well by GREMLIN, given the model’s training on colder cloud tops and areas with lightning present. However, and per an earlier question I had regarding how it might handle lighter precipitation areas, I decided to look at some of the areas of weaker showers near my CWA du jour. For this post, I will focus on the area of lighter MRMS echoes near Madison, WI (near center of image), as well as what appears to be cirrus blowoff being picked up by the radar just to the north.
GRMLN-EFD-4panel image over southern Wisconsin at 2210Z.
The lighter dbz cell right over Madison in these images is still being picked up rather well by GREMLIN, though there do appear to be a couple of lightning strikes associated with it, so that is likely helping. I also noticed some colder cloud tops (what appears to be some cirrus) just to the north of that cell, possibly blowoff from the Madison cell. GREMLIN is picking up on that area, and other apparent cirrus areas as well. However, actual radar tends to pick up these areas as well, especially if the cirrus is thick. That is usually a negative in the operational environment, in that it can make discerning light precip from cirrus difficult. However, for the purposes of evaluating GREMLIN and its ability to mimic actual radar, it could be considered encouraging. Maybe not a mind-blowing observation, but one that struck me as significant enough, especially for light precip or maybe high-based shower situations.
– Loki
