Sitemap Contact Kalender Nieuw Home

Projectendatabank FEDRA





Monitoring Inland and Coastal waters with the APEX Sensor (MICAS)

Onderzoeksproject SR/00/122 (Onderzoeksactie SR)

Personen :

Beschrijving :

Over the past few years, a joint Swiss/Belgian initiative resulted in a project to build a new generation airborne imaging spectrometer, namely APEX (Airborne Prism Experiment) under the ESA funding scheme. APEX is now in its construction phase and will be operationally available to the user community in the second half of 2009. The MICAS (Monitoring Inland and Coastal waters with the APEX Sensor)
project aims at the development of high quality level 3 products related to inland and coastal waters into the APEX processing chain. Hence, the main objectives of the project are:

to find water quality algorithms optimized for APEX which can be implemented into the CDPC and
to automatically generate accurate concentration maps of chlorophyll, CDOM and suspended matter.


MICAS will focus on state-of-the-art algorithms to derive chlorophyll, CDOM and suspended matter concentrations maps. An innovative approach, based on wavelet transform, will be introduced and tested for the inversion of (simplified) bio-optical models. It is expected that this wavelet approach will be less sensitive to disturbing factors such as sensor noise. In the first phase the algorithms will be validated on the basis of simulated APEX spectra from different water types (inland, coastal and estuarine waters). Special attention is given to the simulation of realistic APEX spectra, including atmospheric and sensor noise. An extensive error and sensitivity analysis will be performed to judge the quality of the final products. A validation is foreseen on real APEX images acquired from two distinct water areas: a fresh/inland water lake in Switzerland and the river Scheldt in Belgium. These APEX campaigns will be accompanied by an extensive field campaign during which all necessary data for level-3 product generation and validation will be gathered.


Generally seen there are 4 big steps which can be differentiated:
(i) Initially a database containing SIOPS information from different inland and estuarine water regions will be developed. The SIOPS database will form the basis for the simulation of realistic APEX images of inland water scenes.
(ii) This simulated data will be used for algorithm development and validation of APEX water quality products. Beside known methods like Matrix inversion and Linear Least Square Approach, an innovative approach based on Wavelet based curve fitting will be tested and validated.
(iii) Hyperspectral imagery will be collected with the APEX sensor over 2 test sites, accompanied with extended field measurements to fine-tune the selected algorithm.
(iv) This algorithm will be implemented in the APEX Processing and Archiving Facility to automatically produce water quality products for the end-user community.


The main outcome will be water quality algorithms optimized for the APEX sensor and integrated into the CDPC. These algorithms include new and innovative techniques for handling the inversion problem of bio-optical models fine tuned to the APEX spectral characteristics. As realistic simulations including sensor noise will be used to test the validity of the algorithms for different water types, it is expected that the algorithms will be relatively regional and seasonal robust. This is of large interest to the user community. Not all End Users have the expertise or software to extract water quality data from their images. Without validation, the end user is not interested in pre-calculated products. Therefore validation is an essential part of the project. The user’s community of APEX hyperspectral imagery will not only benefit from the end products, but also from the validated algorithms.
An indication of the sensitivity of the inversion models to errors in the input data.
A successful APEX flight with high quality imagery.
A good quality estimate of the final water quality products is expected

Documentatie :

Over deze website

Uw privacy

© 2017 POD Wetenschapsbeleid