Mapping target domains for flood-tolerant rice varieties in Nigeria using MODIS time series

S.J. Zwart, M Hamdy, O. Adefurin

Research output: Contribution to conferenceAbstractOther research output

92 Downloads (Pure)

Abstract

Rice production in Nigeria occurs mainly in the fertile flood plains of the major rivers and tributaries, areas that are also often affected by floods. Only last year floods destroyed large rice areas in the northern states of Nigeria worth 0,5 M US$ affecting the income of thousands of smallholder farmers and the food security of many more. The Africa Rice Center has developed local rice varieties with the so-called sub1-gene that permits rice plants to remain fully submerged up to two weeks without any yield penalties. To facilitate the efficient introduction of these varieties to farmers who mostly need it, an analysis was made of the frequency of floods during the last 15 years. Floods that do not last more than two weeks were considered and hotspot areas were located.
We used time series of 16-day composite products of MODIS of Normalised Difference Vegetation Index (NDVI) and the Land Surface Water Index (LSWI) between 2001 and 2015. Firstly, the growing season of rice was determined from the NDVI time-series for each of the sub-basins. Then, we evaluated threshold values to mark flooded pixels during the growing season. We summed the flooded periods during each of the growing seasons and we calculated the frequencies for seasons with only one period marked as flooded (flash flood) and those with more than 1 period flooded (stagnant flood). The latter result was validated with reports from floods in the Dartmouth Global Archive of Large Flood Events. The frequency map of flash floods is going to be used by government services to locate rice villages in flash flood hotspots.
Original languageEnglish
Publication statusPublished - 2016
Externally publishedYes
Event11th International Conference of the African Association of Remote Sensing of the Environment 2016 - Kampala, Uganda
Duration: 24 Oct 201628 Oct 2016
Conference number: 11
http://www.aarse2016.org

Conference

Conference11th International Conference of the African Association of Remote Sensing of the Environment 2016
Abbreviated titleAARSE 2016
Country/TerritoryUganda
CityKampala
Period24/10/1628/10/16
Internet address

Fingerprint

Dive into the research topics of 'Mapping target domains for flood-tolerant rice varieties in Nigeria using MODIS time series'. Together they form a unique fingerprint.

Cite this