Notices

1. snow observer ❄️ (cryospheroid@scholar.social)'s status on Wednesday, 18-Aug-2021 16:57:57 CEST snow observer ❄️

   Always get a bit frustrated when people complain about how error-prone precipitation forecasts are. Cloud microphysics is difficult!

   • Pybonacci (pybonacci@mastodon.social)'s status on Wednesday, 18-Aug-2021 16:57:54 CEST Pybonacci

     in reply to

     • Victor Venema

     @cryospheroid @VictorVenema and now put ML/AI into parameterisations... But, in general, there are improvements. Weather modelers are conservative in this sense. We have seen huge improvements during the last decade or so 😃

   • Victor Venema (victorvenema@fediscience.org)'s status on Wednesday, 18-Aug-2021 16:57:57 CEST Victor Venema

     in reply to

     @cryospheroid Even with perfect cloud microphysics and perfect dynamics, I would not expect precipitation forecasts to be particularly accurate more than a day in advance. Especially convective precip.

     Hi, I am Victor. Nowadays I work on the accuracy of climate trends, but as a PhD student I worked on measuring clouds and as postdoc on radiative transfer through clouds. How do you do? Nice to see such a close colleague here.

   • snow observer ❄️ (cryospheroid@scholar.social)'s status on Wednesday, 18-Aug-2021 16:57:57 CEST snow observer ❄️

     in reply to

     • Victor Venema

     @VictorVenema Oh yeah, convection parameterization is a whole other can of worms! (One that I admittedly don't know as much about, it's a bit removed from my area, haha)

     Nice to meet you! I prefer to stay pseudonymous here, so I go by "cryospheroid", but you can also call me A.C. I'm still a PhD student myself; I've done some work with comparing Arctic snowfall remote sensing & reanalyses, and I'm currently working on modelling snow-on-sea-ice (towards improving sea ice thickness estimates).

   • Victor Venema (victorvenema@fediscience.org)'s status on Wednesday, 18-Aug-2021 16:57:57 CEST Victor Venema

     in reply to

     @cryospheroid But even worse than convection parameterizations are people who think it helps to make the resolution a bit higher (but still not resolving turbulence) and just turn it off.

     cough COSMO cough

     Arctic snowfall remote sensing! Every generation the problems get tougher. That is white on white crime.

   • snow observer ❄️ (cryospheroid@scholar.social)'s status on Wednesday, 18-Aug-2021 16:57:57 CEST snow observer ❄️

     in reply to

     • Victor Venema

     @VictorVenema Ah yes, I've heard a bit about those issues, though this is my first time hearing of COSMO specifically. (It seems to be higher-resolution than what I deal with.) That said, I'm doing some work with a model that makes a lot of not-so-physically-robust assumptions myself, so I can't really cast judgement myself, haha.

     The world of Arctic precipitation is fraught with uncertainty! It would be nice to get some more observations, but that's always the case for these things, it seems.