Results
From CrewWiki
(→Validation against A-Train observations) |
(→Comparison of SEVIRI and MODIS cloud retieval algorithms) |
||
| (8 intermediate revisions not shown) | |||
| Line 1: | Line 1: | ||
| - | == | + | == Inter-comparison of SEVIRI cloud retieval algorithms == |
The first workshop CREW-1 concentrated on the inter-comparion of SEVIRI algorithms<br> | The first workshop CREW-1 concentrated on the inter-comparion of SEVIRI algorithms<br> | ||
<center> | <center> | ||
| Line 5: | Line 5: | ||
</center> | </center> | ||
Up to now a detailed analysis was done for cloud mask and cloud top height. In future a detailed analysis of all available parameters is planed. | Up to now a detailed analysis was done for cloud mask and cloud top height. In future a detailed analysis of all available parameters is planed. | ||
| + | |||
| + | == Comparison of SEVIRI and MODIS cloud retieval algorithms == | ||
| + | Cloud properties based on SEVIRI and MODIS observations are compared. The agreements shows the reliability of the algorithms, whereas deviations gives hints for improving the algorithms. The observation from the different satellite platforms differ in viewing geometry, spatial and spectral resolution. The following diagram shows the comparison of cloud optical depth derived from MODIS (MYD06) and SEVIRI (by the [http://www.knmi.nl/ KNMI]). | ||
| + | <center> | ||
| + | [[Image:MODIS_cloud_optical_depth.png|200px|Comparison of MODIS and SEVIRI cloud optical depth]] | ||
| + | </center> | ||
== Validation against A-Train observations == | == Validation against A-Train observations == | ||
| - | Following diagram shows the comparison of the cloud height | + | The signal of the CPR is shown in following diagram. The white symbols show the cloud top height derived by one SEVIRI retrieval algorithm. Colorbars at the bottom show the phase of the cloud. |
| + | <center> | ||
| + | [[File:CTH_validation_CPR.png|400px|Validation of the Cloud top height with CPR measurements]] | ||
| + | </center> | ||
| + | |||
| + | Following diagram shows the comparison of the cloud top height of all algorithms with A-Train measurements. The mean of all SEVIRI cloud retrieval algorithms participating in CREW is shown in black, the measurement of CALIOP in green and the CPR retrieval in red.<br> | ||
<center> | <center> | ||
[[File:CTH_validation_Atrain.png|400px|Validation of the Cloud top height with A-train measurements]] | [[File:CTH_validation_Atrain.png|400px|Validation of the Cloud top height with A-train measurements]] | ||
Latest revision as of 14:37, 1 April 2011
Contents |
Inter-comparison of SEVIRI cloud retieval algorithms
The first workshop CREW-1 concentrated on the inter-comparion of SEVIRI algorithms
Up to now a detailed analysis was done for cloud mask and cloud top height. In future a detailed analysis of all available parameters is planed.
Comparison of SEVIRI and MODIS cloud retieval algorithms
Cloud properties based on SEVIRI and MODIS observations are compared. The agreements shows the reliability of the algorithms, whereas deviations gives hints for improving the algorithms. The observation from the different satellite platforms differ in viewing geometry, spatial and spectral resolution. The following diagram shows the comparison of cloud optical depth derived from MODIS (MYD06) and SEVIRI (by the KNMI).
Validation against A-Train observations
The signal of the CPR is shown in following diagram. The white symbols show the cloud top height derived by one SEVIRI retrieval algorithm. Colorbars at the bottom show the phase of the cloud.
Following diagram shows the comparison of the cloud top height of all algorithms with A-Train measurements. The mean of all SEVIRI cloud retrieval algorithms participating in CREW is shown in black, the measurement of CALIOP in green and the CPR retrieval in red.
Calculation of meaningful Level 3 product
In the third workshop CREW3 in November 2011, another field of investigation is added to the present fields of investigations. We will analyse how to calculate temporal averages of cloud properties (daily and monthly). The effects of different level 2 datasets and of different level 2 to level 3 algorithms will be investigated. In addition the propagation of the uncertainty will be analysed.
