Skip to Main content Skip to Navigation
New interface

Analyse comparative des différents produits satellitaires d'estimation des précipitations en Amazonie brésilienne

Florian Delahaye 1 
1 LETG - Rennes - Littoral, Environnement, Télédétection, Géomatique
LETG - Littoral, Environnement, Télédétection, Géomatique UMR 6554
Abstract : Synoptic and local atmospheric circulations explain the spatiotemporal variability of rainfall in the Brazilian Amazon. This wide geographical area is characterized by a sparse and heterogeneous rain gauges network, which limits an exhaustive monitoring of rainfall. Remote sensing is then the preferred tool to fill this gap space. However, quantitative precipitation estimates (QPE) derived from spectral Infrared (IR) and Microwave (MW), or multi-channel techniques suffer from errors. Sources of uncertainty and bias of QPE may come from the sensor calibration, the algorithms and sampling of rain fields. In this thesis, the evaluation of algorithms TRMM3B42(v6), CMORPH and PERSIANN relies on a rain gauges database which reliability is assured by rigorous quality control. Over the period 2003-2009, the inter comparison of satellite products show low agreement (coefficient of determination < 0.30) with reference database in the Legal brazilian Amazon. TRMM3B42(v6) has the best performance with a correct sampling of rain fields. In addition, a case study shows that satellite QPE errors are larger in the forest, including the product MW MWCOMB. Finally, the satellite grids correction by PDF Matching allows a detailed and improved climatologically analysis compared with reference database
Complete list of metadata
Contributor : ABES STAR :  Contact
Submitted on : Wednesday, May 22, 2013 - 3:27:57 PM
Last modification on : Wednesday, September 28, 2022 - 4:20:10 PM
Long-term archiving on: : Friday, August 23, 2013 - 4:09:49 AM


Version validated by the jury (STAR)


  • HAL Id : tel-00824885, version 1


Florian Delahaye. Analyse comparative des différents produits satellitaires d'estimation des précipitations en Amazonie brésilienne. Géographie. Université Rennes 2, 2013. Français. ⟨NNT : 2013REN20017⟩. ⟨tel-00824885⟩



Record views


Files downloads