Major projects

Developing increased EO capacity for better agriculture and forestry management in Africa


Project location Netherlands
Starting date 01 January 2014
Completion date 31 December 2015
Client HKV
Project leader Dr. J. (Joris) Timmermans
Project officer Bert Boer
Project type Contract research
Budget Euro 43.286


The past years remote sensing data has become more important in water resources management for the Dutch Water boards. However only a small part of the available remote sensing techniques is currently used in this management, such as determining the land use and the estimation of precipitation. These techniques provide the possibility of a completely integrated water resources management system for controlling water levels in the rivers and risk analysis of the water sheds. This has led to the formation of the SAT-Water consortium by several Dutch water boards. The goal of this consortium is the combined purchase of strategic and operational data and hydrological information for water management. This will be achieved by focusing on

  1. Building up carrying capacity by joint use by the water boards of new methods and instruments
  2. Reduce burocratic loads on the water boards by joint-purchase of data
  3. To reduce inefficiency by coupling knowledge, network and research questions;
  4. Efficient interaction with the suppliers of the data
  5. Anticipate towards the development of precision agriculture by web based services to provide information to the farmers.

In this SAT-Water frame work it was deemed necessary not only to have data on precipitation (by satellite observations), storage (by ground measurements)  but also estimations of the evapotranspiration. Within the water cycle evapotranspiration provides the largest sink of water. Therefore in an integrated water resources management system it is of vital importance before any water balance can be performed. As a consequence the SAT-data project was launched, which is executed by a consortium consisting of the ‘HKV in line with water’ consulting company and the University of Twente (water resources department at the faculty of ITC).

Product requirements

Within this project an evapotranspiration product over the Netherlands is developed that provides accurate estimations at high spatial resolution (250m x 250m) under clear-sky and clouded conditions. The boundary conditions of the product should be

  • Open standards. In a world where more open source and open data (instead of proprietary software) are becoming the standard, algorithms that are necessary for management systems should be made readily available to the user and public.
  • Physical modeling. The method to determine evapotranspiration should be as physically based as possible to circumvent miscalibration of empirical parameters. In addition the model should provide consistent estimations of evapotranspiration under both clouded and clear-sky conditions. 
  • Risk analysis: The methodology should be robust and be able to switch between data delivery platforms. When using remote sensing within an operational water management system there is always a large risk concerning breakdown of satellites. The methodology employed should therefore be able to switch between satellite systems while reducing downtime.


Considering the requirements of the product a distinction was made between cloudy and clear-sky conditions.

  • For clear-sky conditions the SEBS algorithm is chosen as the ideal baseline algorithm. This model provides the ideal balance between the amount of representativeness of the physical parameterizations, the amount of inputs required by the model, and the computation speed of the algorithm. In addition SEBS has been validated in several studies and is still being further developed by the University of Twente. However SEBS does not provide evapotranspiration estimations under cloudy conditions.
    • To drive the model with remote sensing data, observations by the MSG (Second Generation Meteosat) Satellite are used.  This satellite provides near-real time observations of albedo, LAI, LST. The advantage of this satellite is its high temporal frequency, but is limited by the coarse spatial resolution. The coarse resolution of this satellite is increased by disaggregation using MODIS observations to high resolution (to 250meter). A second disaggregation step is then used (using Formosat and Radarsat-2 data) to scale to 8m resolution.
  • During cloudy conditions a weather model (Hirlam/Harmonie) is employed to provide realistic estimations of the land surface processes. These data can then be used to estimate evapotranspiration during overcast days. In addition the weather model provides hourly observations of meteorological values required for the SEBS algorithm.
  1. Home »
  2. Project services »