Radiation / Clouds
Based on satellite images the current cloud situation is observed and the global horizontal irradiation (GHI) can be derived precisely. Therefore this satellite derived radiation is very useful to estimate the yield of solar energy application. While historical values are used to asses the long term production, real time values allow nowcasting and monitoring of the actual system behaviour. Shortwave radiation is the major source of energy for all plants, so these datasets are interesting to environmental and agricultural purposes, too.
Presently meteoblue offers the following datasets:
- Meteosat Cloud Physical Properties (KNMI) - covers Europe, the Arabian peninsula, Africa, Brazil, and parts of its neighbouring countries
- Other satellites will be implemented soon (North & South America, India, Japan etc.)
Meteosat Cloud Physical Properties (KNMI)
| Field | Metadata |
|---|---|
| Title | Meteosat Cloud Physical Properties |
| Short Title | Meteosat, MSGCPP |
| Version | 4.0, 2011 |
| Theme | meteosat, shortwave radiation, cloud properties, satellite |
| Period of Time | 2014 to present |
| Frequency | daily |
| Update timelag | < 24 hours |
| Time-resolution | hourly |
| Spatial Type | Irregular grid, Geostationary Satellite at 0.0N / 0.0S |
| Spatial extend | 50°W to 50°E / 80°S to 80°N |
| Spatial resolution | 12 km at Nadir point (0 North, 0 East) |
| Coordinate system | WGS-84 |
| Earth model | WGS-84 |
| Spatial reference system | |
| Publisher | KNMI (Dutch Meteorological Institute) |
| Date First Published | December 2011 |
| Date Obsolete | |
| Description | Indirectly derived/modelled shortwave radiation and cloud properties. The approach requires two independent steps: 1. Cloud properties are determined from narrow-band satellite radiances. 2. These cloud properties are used together with data on water vapour column and surface albedo to calculate the atmospheric flux transmittance. The retrieved irradiance has an underestimation bias by about 3-4 W/m2 throughout the year, corresponding to an underestimate in atmospheric flux transmittance of about 0.015 in summer and 0.04 in winter. From a least-squares linear regression, residual standard deviations of 56 W/m2 (0.072, 17.0%), 11 W/m2 (0.052, 10.8%), and 4 W/m2 (0.021, 4.2%) are found for hourly, daily and monthly mean irradiance (transmittance, relative error), respectively. These findings indicate that the accuracy of the retrieval is comparable to first-class pyranometers in the summer half year (5% of daily-mean values), but significantly lower in winter. Two aspects requiring further investigation have been identified: 1. For thin clouds, the atmospheric flux transmittance is strongly underestimated. 2. The retrieval accuracy is reduced over snow-covered surfaces. |
| Detailed specification | HM Deneke, AJ Feijt, RA Roebeling. Estimating surface solar irradiance from METEOSAT SEVIRI-derived cloud properties, published, Remote Sens. Environ., 2008, 112 |
| Homepage URL | http://msgcpp.knmi.nl/mediawiki/index.php/MSGCPP_product_description |
| Security Classification | public |
| Access Level | No constraints |
| License | http://msgcpp.knmi.nl/mediawiki/index.php/MSGCPP_product_description#Access_constraints |
Current dataset issues
The Meteosat data is available from one hour after sunrise until one hour before sunset, which means for the first and last hours with low sun angles, the radiation values are equal to zero. This leads to low values when data is aggregated to daily or monthly sums, with an underestimation of about 10 %. Thus we recommend to replace all values below 50 W/m2 with radiation values from a weather model (e.g. NEMSGLOBAL).