tag:blogger.com,1999:blog-7729093380675162051.post6555865811832627332..comments2024-03-28T13:56:47.604+11:00Comments on moyhu: Evenly spaced icosahedral grid for global temperatureNick Stokeshttp://www.blogger.com/profile/06377413236983002873noreply@blogger.comBlogger6125tag:blogger.com,1999:blog-7729093380675162051.post-20238076381239392512019-04-03T01:14:40.112+11:002019-04-03T01:14:40.112+11:00The NullSchool type wind depiction can be very mis...The NullSchool type wind depiction can be very misleading, especially in regard to actual air movement, but I find it helpful for visualizing convergence and divergence in wind fields and when overlayed on shaded temperature it helps for visualizing cold and warm air advection. It should be possible to create a dynamic wind animation visualization over time (instead of for a static time) that would more accurately represent the wind movement, but I have yet to see one using real-time forecast output. The time steps might need to be fairly short (sub hourly) to get a smooth animation.<br /><br />Many years ago I wrote a program to plot surface wind and air quality data along with air parcel trajectories and plume simulations based on hourly vector average wind measurements. I found very quickly that hourly wind vectors plotted to scale look very tiny when viewed at large scales, but most of my work was for urban and regional scales where it worked quite well for visualizing air motion relative to air pollution.Bryan - oz4casterhttps://www.blogger.com/profile/18027990322659101002noreply@blogger.comtag:blogger.com,1999:blog-7729093380675162051.post-63714727849263541982019-04-02T13:50:57.948+11:002019-04-02T13:50:57.948+11:00"The wind depiction reminds me..."
Yes, ...<i>"The wind depiction reminds me..."</i><br />Yes, it's similar, and not quite what it seems. It's just a visualisation attached to a static wind field. It shows how air would move if the wind remained constant, which of course it doesn't. We used to make visualisations like this in CFD before we could dynamically solve the Navier-Stokes equations.<br />Nick Stokeshttps://www.blogger.com/profile/06377413236983002873noreply@blogger.comtag:blogger.com,1999:blog-7729093380675162051.post-30601195243594756902019-04-02T13:44:30.779+11:002019-04-02T13:44:30.779+11:00Nick, thanks for the link to the detailed ICON pap...Nick, thanks for the link to the detailed ICON paper and it's impressive that you get 1.05.<br /><br />Olof, thanks for the VentuSky link. I have never seen a depiction of forecast lightening like they have. The wind depiction reminds me of <a href="https://earth.nullschool.net/#current/wind/surface/level/overlay=temp/orthographic=13.04,45.80,414" rel="nofollow">NullSchool</a>, but I don't see an option for orthographic projection. However, the interface is very easy to use and intuitive.Bryan - oz4casterhttps://www.blogger.com/profile/18027990322659101002noreply@blogger.comtag:blogger.com,1999:blog-7729093380675162051.post-75474890814159107962019-04-01T19:37:28.774+11:002019-04-01T19:37:28.774+11:00If anyone wants to see the ICON model in action, t...If anyone wants to see the ICON model in action, there is a nice app <a href="https://www.ventusky.com/" rel="nofollow">here</a><br />You can choose between the ICON and GFS weather model..Olof Rhttps://www.blogger.com/profile/18244733455655978307noreply@blogger.comtag:blogger.com,1999:blog-7729093380675162051.post-37227508863756465962019-03-31T14:20:03.111+11:002019-03-31T14:20:03.111+11:00Thanks, Bryan
No, I hadn't seen that. I knew G...Thanks, Bryan<br />No, I hadn't seen that. I knew GFDL was using a <a href="https://www.gfdl.noaa.gov/fv3/" rel="nofollow">cubed sphere</a>, which is similar. I don't think that anyone should be using lat/lon for modelling work nowadays.I see <a href="https://rmets.onlinelibrary.wiley.com/doi/10.1002/qj.2378" rel="nofollow">here</a> that they get an area ratio ranging from 1.38:1 to 1.53:1. It may be good enough for what they need, but the simple transformation described here gets it down to about 1.05:1.<br />Nick Stokeshttps://www.blogger.com/profile/06377413236983002873noreply@blogger.comtag:blogger.com,1999:blog-7729093380675162051.post-64382333521564534422019-03-31T07:02:45.445+11:002019-03-31T07:02:45.445+11:00Nice work Nick. Interesting results. Have you lo...Nice work Nick. Interesting results. Have you looked at the DWD ICON weather model (<a href="https://www.dwd.de/EN/research/weatherforecasting/num_modelling/01_num_weather_prediction_modells/icon_description.html" rel="nofollow">here</a>)? It uses a global icosahedral grid at about 13 km resolution. The ICON output can be seen <a href="https://www.tropicaltidbits.com/analysis/models/?model=icon&region=aus&pkg=mslp_pcpn_frzn&runtime=2019033012&fh=6" rel="nofollow">here</a>.<br /><br />I would guess that the ICON surface air temperature initial condition grid could be analyzed in similar fashion to the reanalyses performed with the GFS and ECMWF weather models, although I have not seen any such effort so far. If it were available, it would be interesting to compare with your effort.Bryan - oz4casterhttps://www.blogger.com/profile/18027990322659101002noreply@blogger.com