If you have Radarscope, that is best. If not:
local radar: https://weather.cod.edu/satrad/nexrad/?parms=TLH-N0B-0-24-100-usa-rad
mosaic radar: https://weather.cod.edu/satrad/?parms=local-FL_Panhandle-comp_radar-24-0-100-1&checked=map&colorbar=undefined
Until a well defined center of circulation forms (generally when it becomes a Tropical Storm) we won’t get accurate models. The variation in the HAFS-B from 12z to 18z is exactly what to expect until then.
0.1" here too, and yesterday models were predicting closer to an inch by now. Further overnight storms are way less now… we need rain. I ended up watering some of my plants mid day today.
You can see the CDO is now very much displaced from the CoC, which is now towards the eastern edge of the CDO. Shear is having a major impact. I find it interesting that NHC is unsure what shear will be in 24 hours, with ECMWF increasing shear and GFS decreasing shear.
Looks like the eye is trying to clear again, last couple times it tried and failed, so we’ll see if it succeeds this time. One big difference I’m seeing on satellite since earlier today is there are evident gravity waves again; haven’t seen those in over 24 hours since sheer started affecting Beryl.
Most modern models are not built on statistical analysis unlike a lot of models 20 years ago. Most models today are fluid dynamic models that attempt to simulate the atmosphere. Errors are due to limited and erroneous data ingested (which is why you often see NOAA missions dropping dropsondes in front of hurricanes), granularity of the model (models grid the planet and each "pixel" is modeled at depth), and simplifications of the fluid dynamic equations to allow modern supercomputers to produce a result in a reasonable length of time (this is a big part of why you see discrepancies between models). Global models are typically worse at hurricanes due to lower granularity resulting in the core pressure being unable to be modeled accurately; this error tends to be amplified in deeper, stronger hurricanes. Hurricane models are typically lagged by 6 hours because they are typically based off a 6 hour earlier global model (also why you need to wait 12 hours after formation of a tropical storm before you can trust model output).