@kameshmanda

If there is a weather station nearby, data can be retrieved from infoclimate.fr (this link shows the recorded data for Vienna).

For GFS data there is an archive on http://rda.ucar.edu/data/ds084.1. You can register there for free, but you have to write a program to extract the data from the grib files.

@Yves70

Thanks for your additions. I think that some basic synoptic features are useful for a wide group of windy customers. Maps with very special parameters for people with a deep understanding of meteorology and thermodynamics would be a waste of (development) time.

In my opinion, the possibility to combine maps from different pressure levels allows to evaluate the weather situation on a synoptic scale and is a good balance between costs and benefits.

The following example combines precipitation with 500hPa geopotential (made with Photoshop) and shows what I mean. Now we can see the cause of the predicted surface weather in the upper air:

I don't know if synoptic meteorology is your focus, but if you also want to address this, I have some suggestions. With your zoomable maps you already have a code base that is almost nowhere to be found.

Isolines:

• The most important information on classic maps or weather sites. If the thickness and the colour (white/black) is adjustable, they would be more visible and more useful.
• At the moment the distance between the lines ist arbitrary. A distance of 40m (geopotential height) with values divisible by 40 is common.
• More effort, but extremely useful: Decouple the isolines from the selected height of the map. For example, the 850hPa temperature and the 500hPa geopotential can be displayed. This feature is not available anywhere else, because the maps are mostly server rendered. So you would be the only ones.

Charts:

• The data of the selected map (e.g. 850hPa temperature) should also be viewable as a diagram using the location picker.

Colors:

• Make it possible to disable color gradients between each value. The map may not look as nice, but temperature gradients for example will be more visible.

Kind regards
Michael, Austria

First of all: thank you for offering this great service for free. It is more than our hole national weather service in Austria can provide!

To close the gap between the “classical” representation of model data and your site I think a 2nd layer, which visualizes the numerical data as a contour plot, would be a general solution. With an independent z layer every user can combine 2 arbitrary model values (for example 850hpa temperature as colours and the sea level pressure as contour lines).

A second improvement would be the possibility to disable the color gradients. So it’s easier to detect cold fronts on the 850hpa temp map.

Thanks and keep on your great work!