Informations et ressources scientifiques
sur le développement des zones arides et semi-arides

Accueil du site → Master → Pays Bas → 2021 → Use of a DEM and NDVI time series to map and monitor horticultural areas in Tigray, as input for a better-informed food supply chain.

University of Twente (2021)

Use of a DEM and NDVI time series to map and monitor horticultural areas in Tigray, as input for a better-informed food supply chain.

Molla, Kisanet Haile

Titre : Use of a DEM and NDVI time series to map and monitor horticultural areas in Tigray, as input for a better-informed food supply chain.

Auteur : Molla, Kisanet Haile

Etablissement de soutenance : University of Twente

Grade : Master of Science in Geo-Information Science and Earth Observation 2021

Résumé
This paper identifies, maps, calculates the percentage, and monitors the presence of horticultural crops within small clusters of fields in Tigray. First NDVI time series were used to select the possible are of interest (AOI). The first approach was with the perception that higher NDVI values during the dry season can possibly represent horticultural areas and can be used to square the desired results. But the first approach has later proven that it is not possible to identify horticultural areas based on NDVI only. Hence, DEM with additional selection parameters was used, to get the possible area of interstates. After having those desired areas of interest, the NDVI time series were used to Further limit potential areas of interest and lead to the final AOI’s. This was done by using the unsupervised classification of the extracted areas from DEM and the Pre-Processed NDVI-data (DN values) image from 20 years Spot-VGT + PROBA-V NDVI-data. After acquiring the final AOIs the percentage of each AOIs that is covered by horticulture was also calculated. This was done by pixel level, by creating random points of 1km pixels for each specific NDVI-class in ArcMap and checking how much percentage of that pixel is covered by horticulture on google earth image. Resulting in most of the classified classes containing less than 40% of horticulture. This was followed by monitoring for the selected clusters of fields that containing greater than 40% of horticulture. This was done by assessing their temporal NDVI-patterns, which periods crops were grown, and what their performance was. All final AOIs show higher NDVI values during Kiremit, which is the rain season of the region. In most of the classes that have a higher percentage of availability of horticulture, their NDVI values during the dry season are relatively higher. This is done for the year (covering 2010 to 2019).

Présentation

Page publiée le 18 décembre 2021