Operational evapotranspiration estimations for hydrological applications using MSG and Envisat data

-

To establish operational MSG-receiving stations at three locations (VU/IVM, ITC and FAO-Artemis), to exchange experience and to collaborate in finding solutions for data storage and processing.

-

To adapt the codes of the published evapotranspiration models DEMI-SEC (Peters) and ESTA (Timmermans et al) for use with MSG. The rationale for using two independent evapotranspiration models is the possibility to cross check results for debugging purposes without having to do a full validation (which should be done, but in a follow up validation project).

-

Since there is no standard atmospheric correction for MSG visible and NIR observations, higher resolution standard products from other satellites and existing ground validation stations will be used to verify and tune existing approaches (Peters, 1995) to atmospheric correction of MSG. For this purpose EVAPO-MSG will concentrate on ENVISAT standard products (surface temperature from AATSR, albedo and NDVI from MERIS).

-

To do a test run of both evapotranspiration models for a selected period for two selected catchments.

-

To demonstrate to the end-user FAO the applicability of the evapotranspiration calculations for hydrological applications by using the ET results in a calibrated and well documented hydrological model (Stream).

-

Set-up of receiving dish, software and front-end + back-end computers, test of processing performance etc., protocols for MSG receipt + storage of data, for three operational receiving stations.

-

Adjustment/recalibration of DEMI-SEC and ESTA model code to use on MSG data using outputs from studies Peters 1995, Timmermans 2003, existing user requirements (e.g. ACMP project). Implementation of split-window atmospheric correction for TIR bands, albedo estimation upgrade, NDVI implementation etc.

-

Verification of MSG albedo, surface temperature etc. using ENVISAT standard products through implementation of radiometric routines, atmospheric routines combined with upscaling of ground-based measurements and upscaling of ENVISAT products.

-

WP2: MSG receiving stations (at ITC, VU/IVM and FAO: only advice (by means of writing a protocol) or full installation – operational system. Role of FAO as they are currently working on installation and training. Discussion on status with Peters of FAO and VU/IVM. Additional remark: for use of ITC developed data retrieval software a disclaimer should be prepared and signed by the project partners, provided that they have an agreement with Eumetsat. When using the software acknowledgement should be given to ITC as well.

-

WP3: Appropriate definition of input, activities and output (also applies to WP2 and 4), indicating the type of atmospheric correction method(s) (are these rough empirical or highly advanced transfer functions!). Output as seen by us: import of MSG data, atmospheric corrections (SMAC for the visible and method described by Peters for the thermal, requires additional data like water vapour, ozone, aerosol: who is collecting this? and from which data bases). Use of atmospheric corrected data in ESTA by Timmermans, requires adaptation of his model code. Output ET maps of time series (timing, time interval and area to be determined, e.g. Zambezi catchment, therefore additional atmospheric data needs to be collected from the equator to 25 degree South latitude) and draft description of the approach followed. This description will be further elaborated into a peer reviewed article.

-

WP4: Atmospherically and geometrically corrected SEVIRI images, transformed into temperature, albedo and NDVI are compared with Envisat standard products. Procedure for upscaling needs to be addressed. Output of WP3 and WP 4 results in a document describing the procedure(s) and discussion on validation using Envistat standard product(s), apart from the data itself which are needed for the other project partners.