I would like to see Windy desktop add the capability to split screen into two windows side by side, each displaying a different layer but for the same time, location and magnification.

For example, weather radar on the left and satellite IR on the right, as shown below.

Please bring back 2hr satellite and get rid of forecast radar (which is so inaccurate it is useless) and make radar 2hr to match satellite.

I would like to suggest that Weather Radar playback be extended to 2Hr from the current 1Hr duration to match the duration of current satellite image playback (all 2Hrs). This makes it easier to study correlation of precipitation patterns with cloud patterns.

Please update reported temperatures/winds whenever whatever layer that is on display is updated.

Currently, reported temperatures shown on Windy stay unchanged and I have to manually refresh the Windy window to get them updated, or change to displaying reported winds and then back. Many times I forgot to manually update them before saving the satellite images and wind maps the next day, which spoiled the saved images.

For instance, if I leave the orange colored circle on the right end of the horizontal time slider slot at the bottom of satellite image layer or weather radar layer, the displayed image will update itself whenever new data is available, but the reported temperatures/winds never do. I would like the latter to update as well.

At 600hPa there're gray shading boxes indicating cloudiness (my best guess). But the dew point is 23 degC below atmospheric temperature.

The new satellite layer released on web app today introduced a major change to the time slider. Instead of stopping every 15 minutes while in 6h mode, now it stops every 30 minutes in response to LEFT/RIGHT Arrow keys. Question: are all the other minutes in between not real satellite image data but computer generated?

There is no more 6h and 12h modes. Now, everything older than 6h is stepped by 1hr instead of 1/2 hr by LEFT/RIGHT Arrow keys. Question: Are we getting real satellite image hourly and the rest are all computer generated?

Th

@hangInThere
Here's one that has an opposite problem--where (~475hPa) the dew point is very close to atmospheric temperature it's not marked out as highly cloudy--in addition to the aforementioned problem :

NoData in Radar over China in Windy.

But all is good at China's National Meteorololgical Administration fused all-China radar map.

And their individual city station radar maps too.

@strmchaasr @Filip_K
I would add that ice water path (IWP) is critical indicator of huge lasting rains in the tropics. South Korea's GK2A geostationary weather satellite captured images and derived products are provided on Korea Meteorological Administration (KMA)'s website: https://nmsc.kma.go.kr/enhome/html/satellite/viewer/selectSatViewerEnhome.do?dataType=gk2a#

Below screenshot shows the huge ice clouds over South China Sea just off Vietnam and off Malaysia. Next screenshot shows Windy's weather radar layer over SCS. In comparison, there isn't much in Liquid Water Path (LWP).

Ice Water Path (IWP):

Weather Radar layer (Windy):

Liquid Water Path (LWP):

@idefix37 I see. So what Mark Ingalls wants is forecast precipitable water. What I want is real observed precipitable water, a derived product from satellite observation. See https://cw3e.ucsd.edu/satellite/

@Ondřej-Šutera
Will the developer look into reproducing the TCON error the the 1 April 2023 noon radiosonde graph for Singapore Changi Airport to find and correct the error at the root?

What is the current 12hr satellite image time resolution? I can get 10 mins resolution from Korea meteorological department for up to 18hrs to look at East Asia and Southeast Asia. I remember Windy's 12hr satellite image sequence was rather jerky with animation turned off. I turn off animation in order to examine cloud flows closely. This becomes very difficult to do on the low altitude clouds when they are overlapped by patches of higher altitude clouds and they can only be seen in the gaps between the latter but their motions become impossible to discern when the time resolution is much more than 10 mins. So before I pay for the premium subscription I would like to know what I will get.
Another question, what is the difference in subscriber obligation btwn the one-time purchase and the annual subscription? Can one cancel the annual subscription after say two years?
BTW, this is a great service, and I believe in paying for a good service that I use everyday to derive benefits from, especially for the unbelievably low price that it is charging. Just my personal opinion. Other users can have their own.

I would like to request adding a button to save snapshots of all altitudes of the wind maps to files named after a user given prefix and timestamp and altitude label. I do this several times a day in order to manually archive the altitude wind maps besides video recordings of the satellite images and weather radar images. It's very, very time consuming.

In this Radiosonde skew-T chart for Singapore, the only altitude range 400hPa to 300hPa that has low relative humidity is marked as cloudy whereas those other altitude ranges actually have very high RH but not marked as such.

Currently, the IR satellite images only colorize the high altitude clouds while giving grayscale to mid and low altitude clouds. I would suggest adding a completely new layer for cloud top pressure to cover the entire range of hPa values from 100 to 1050 and adopt the color scale built from purple, black, blue, green, brown, and red and within each color apply grayscale, like what the Korea Meteological Agency does in its Image Viewer (https://nmsc.kma.go.kr/enhome/html/satellite/viewer/selectSatViewerEnhome.do?dataType=gk2a# ). With this layer, one can see all levels of clouds at once. For example, this past afternoon there were simultaneously high & mid altitude clouds from the west and low altitude clouds from the east over mid Sumatera Island, but none of the Blue/Visible/IR satellite image layers can distinguish the mid from the low altitude clouds, whereas the KMA's Cloud Top Pressure product does it so well. Go see it for yourself. Set the time to 1020 UTC and duration to "Last 12 Hrs". Try it. You will find that in such complex multi-layer cloud flow situation their color codes perform really well.

@idefix37
Your suggestion to show historical positions in markers with time and linked by lines would be very helpful.

I would add a further suggestion that, upon clicking on a ship marker, ALL its historical positions are brought up in markers--without these historical markers, the historical weather measurements in chart become meaningless for ships.

A further suggestion is to reposition the vertical hairline on the chart to the time that correspond to a historical ship marker that has just been left clicked at.

The above would be very helpful where one needs to determine where & when a ship encountered a particularly challenging weather situation, e.g. like what happened in the south of the Strait Of Melaka this morning, with strong winds at 48km/hr. This is essential information to further investigate what weather phenomenon brought this strong wind. I hope Windy R&D team can implement these improvements. The weather data from ships are too crucial to be lost--for the study of adverse weather events at sea one can learn how to better forecast using historical data. See the adverse event in chart below:

@hangInThere
Okay, this is what I do for now:

[[0,[24,24,24,255]],
[38.67625899280576,[240,240,240,255]],
[40.51798561151079,[229,172,255,255]],
[53.410071942446045,[183,95,226,255]],
[79.19424460431655,[149,38,205,255]],
[81.03597122302158,[64,64,163,255]],
[106.82014388489209,[70,106,227,255]],
[130.76258992805757,[49,153,231,255]],
[154.70503597122303,[41,187,246,255]],
[186,[163,253,61,255]],
[195,[237,208,59,255]],
[205,[251,128,34,255]],
[214,[210,49,4,255]],
[223,[122,4,3,255]],
[256,[48,0,0,255]]]

which results in this color scale:

and the IR image:

Instead of cloud top pressure, the color scale goes by cloud top temperature in Kelvins.

Here, ground temperature is around 25 degC, i.e. 298K, and assuming 6.5K drop in temperature for every 1km increase of altitude, low clouds with top below 1km should read higher than about 291K, thus colored in the lightest purple or even white.

The temperatures in the color scale need to be adjusted by however much the ground temperature is higher or lower than 25 degC. Obviously if the ground is 40 degC like in Europe's summer 2022 (15 degrees on top of 25 degC) we will be making an error in cloud altitude of more than 2km and mistaken mid altitude clouds for low altitude if the color scale's temperatures aren't adjusted accordingly.

If someone can improve further, please post a reply below.

@Filip_K That's fast! I just screenshot Singapore's radar image for the same 9PM and showing rain area that matches MetMY much better.

@Ondřej-Šutera
One more suggestion: instead of the user having to place the spot where the lifted dry adiabatic parcel temperature reaches its dew point, place it automatically and draw the dry & moist adiabatic curves thru this spot automatically. Saves all the trouble for the user.