Logo Tropentag

Tropentag, September 14 - 16, 2022, Prague

"Can agroecological farming feed the world? Farmers' and academia's views."


Reliability of satellite-based rainfall products for water management studies: a case study in the Ankavia river basin in Madagascar

Ramahaimandimby Zonirina1, Randriamaherisoa Alain2, Charles Bielders1

1Université Catholique de Louvain, Earth and Life Institute, Belgium
2Université d'Antananarivo, Génie Civil, Madagascar


Abstract


Hydrological modelling for water management in large watersheds requires accurate spatially-distributed rainfall time series. In case of low coverage density of ground-based measurements, satellite precipitation products (SPP) constitute an attractive alternative, the quality of which must nevertheless be assessed. The objective of this study was to evaluate at daily/event/hourly time scales the performance of six SPPs (African Rainfall Climatology Arc v2, Climate Hazards groups Infra-Red Precipitation with station data CHIRPS, ERA-5, Integrated Multi-satellitE Retrievals for Global Precipitation Measurement GPM-IMERG, Precipitation Estimation from Remotely Sensed Information using Artificial Neural Networks- Cloud Classification System PERSIANN-CCS, African Rainfall Estimation REF v2) against a 2-year record from a network of 14 rainfall gauges located in the Ankavia catchment as part of the GIRE-SAVA project (Madagascar). The results suggest that GPM-IMERG outperforms other SPPs on a daily scale (R2=0.63 and RMSE = 12 mm/day). All SPPs, except ERA-5, overestimate the no/light rain class value. Arc v2, GPM-IMERG, PERSIANN-CCS, REF v2 all underestimate moderate rain, whilst CHIRPS, ERA-5 overestimate it, and only GPM-IMERG, PERSIANN-CCS could estimate violent rain satisfactorily. According to the categorical statistical criteria, GPM-IMERG appears to detect no/light/moderate rain events relatively well, whereas PERSIANN-CCS outperforms GPM-IMERG for heavy rain events. Only the GPM-IMERG was evaluated for event and hourly time scales. At the event time scale, the Probability Density Function (PDF) value shows good agreement, however, at the hourly scale, the quality is worsening (R2=0.22). Overall, notwithstanding their limitations, GPM-IMERG products can be regarded as a reliable precipitation source for hydrological modelling for Ankavia watershed.


Keywords: Ankavia catchment, GIRE SAVA, GPM-IMERG, madagascar, satellite precipitation products


Contact Address: Ramahaimandimby Zonirina, Université Catholique de Louvain, Earth and Life Institute, Rue des sports 11/303, 1348 Louvain-la-neuve, Belgium, e-mail: zoramahaimandimby@gmail.com


Valid HTML 3.2!