340 Version of SIS verification post

340 Version of SIS verification post

by Theodoros Gkousarov -
Number of replies: 3

Hi All,

Today I downloaded the newest version of SIS (340) daily means for the first 6 days of June 2013. I compared the accumulations over these 6 days ( in converted to MJ/m2) against the total surface radiation observations from 8 stations over the UK, (Met Office netowork).

Figure 1. Total SIS accumulations 01-06June (0.2 degrees resolution)red fonts, station totals for the same period black bigger fonts. 

 

We can easily observe an overestimation of SIS by at least 10%. I know that the period is very sort to gain any conclusions and also the station observations are site-spesific while the satellite estimate is averaged over the whole grid square, but  would like to see some results here if possible from your verifications. 

This results are preliminary and I will wait at least to have the whole months of June to run a more thorough analysis.

 

Kind regards,

Theo

In reply to Theodoros Gkousarov

Re: 340 Version of SIS verification post

by Theodoros Gkousarov -

Please ignore the previous figure as I spoted a mistake in my calculations.

Today I calculated the totals for the first 10days of June and compared it against the total accumulations from 8 station based measurements. The results are IMPRESSING!

The difference between the 7 stations is less than 2% apart from the station in cambridge which is constantly recording lower values and is around 20%. 

The 7 stations are operated by the Met Office, while the station in Cambridge is from another operator ( dont remember now but will look it up). This will give me grounds to send an e-mail to the operator and let them know about this.

 

Figure 1. Total accumulations of SIS (MJ/m2) for the first 10 days of June 2013. Blue triangles indicate the locations of surface based stations.

 

I am very excited, but I wonder whether there is any chance to get the EDR in much higher resolution? Currently I am using the WUI to make it from Sinusidal to regular lat-lon grid 0.2degrees. Is there any chance we could have the EDR in 0.03 degrees?

 

Thank you so much,

Theo

In reply to Theodoros Gkousarov

Re: 340 Version of SIS verification post

by Jörg Trentmann -

Dear Theo,

thanks a lot for sharing your results on the comparison between the surface measurements of the global radiation and the CM SAF SIS data for June 2013. We currently cannot (due to limited ressources) monitor the performance of the satellite data in an operational manner, so your results are very useful for us! We are glad to see that removing the software bug indeed improves the results (as it should, but one never knows.....).

At this point, the original spatial resolution of the CM SAF data on the sinusoidal grid is 15 x 15 km. Of course you can grep the data in the original resolution, but the spatial information should be very similar to the 0.2 deg you are using already.

You can get access to the instantaneous data at higher spatial resolution, namely at the satellite sensor resolution, i.e,, approx. 0.05 deg. However, you would have to take care about the temporal averaging, if required. For access the instantaneous satellite data (i.e., hourly) you might have to modify your customer class in your user profile.

Since we often receive the request for higher spatial resolution of our operational data, we do consider to provide the data at higher spatial resolution in the future.

Thanks and kind regards,
Jörg

In reply to Jörg Trentmann

Re: 340 Version of SIS verification post

by Theodoros Gkousarov -

Thank you very much Jorg,

I just modified my customer class and hopefully to get the access to hourly soon.  Then I can also use the instantaneous data to verify against instantaneous land based observations ( just yesterday got 5 min intervals data from a private high quality wx station).

I will keep CMSAF users informed with my results here in the forum. 

kind regards,

Theo