University of Twente Student Theses


Mapping crop field probabilities using hyper temporal and multi spatial remote sensing in a fragmented landscape of Ethiopia

Mohammed, Issamaldin Mohammed Alshiekh (2019) Mapping crop field probabilities using hyper temporal and multi spatial remote sensing in a fragmented landscape of Ethiopia.

[img] PDF
Abstract:Crop production is crucial informationfor food security analysis.Crop productionisdefinedasa product of crop area(CA)and yield. Therefore,crop area should be estimated accurately to obtain reliable crop production information.Agriculturalcensuscontains accurate information about CA but itis costly, and it lacks appropriate temporal and spatial information for reliablefrequent crop area estimate. Hyper-temporal remote sensing can capture the general agro-climatic conditions butit is too coarse spatially to capture variability in CA over fragmented landscapes. High-resolution remote sensing can capture the variability of CA butit can not capturethe climatic conditions due toits low temporal resolutionand subsequently fewer images may be available(i.e. because of persistentcloud cover during crop growing seasons).SPOT-VGT NDVI series (1999-2017) was used to identify agro-ecologicalzones through ISO-DATA unsupervised classification. Then these zones were integrated with reported crop area statistics through stepwise linear regression to produce coarse field fractions (1km-resolution). Landsat-8 images (2013-2017) were used to extract moderate resolution (30m) long-term averagedry and wetseasonsNDVI per each agroecological zone.DryandwetseasonsNDVI, elevation,slope, and 1km field fractionswere incorporatedinageneralisedadditive model (GAM). Through the GoogleEarth platform, 271 frames (30mx30m) were visually interpreted to estimate field fractions of these frames for model calibration and validation. The overall deviance explained by the model was62%. The 1km field fraction was found to be the most importantpredictor in our model as it explained 24% of the deviance. As many researchers focus on wet season NDVI, our results showed that the dry season NDVIwas the second importantpredictor and explained 16% of model deviance. Elevation added more explanatory power to themodel (i.e. explained 15% of the deviance). The field fractions predictions (30m-resolution) produced by our final global model explained 77% of the variation in 81 actual fractions observations.To demonstrate the capabilities of the developed global GAM(i.e. over whole Oromia region), a localisedGAM was developedwithin one agroecological zone andthen the global GAM and local GAM were evaluated with an independent test set. The global model performed closelyto the localmodel. These resultssupportsthat hyper-temporal remote sensingcan be effectivein addressing the climatological differences regarding CA estimation. The method can be appliedby governments and researchers for further studies and to aid in decision making regarding cropping and food security policies. Future work should consider involvingadditional predictors to the GAM such as:socio-economic variables, other vegetation indices, andradar images.
Item Type:Essay (Master)
Faculty:ITC: Faculty of Geo-information Science and Earth Observation
Programme:Geoinformation Science and Earth Observation MSc (75014)
Link to this item:
Export this item as:BibTeX
HTML Citation
Reference Manager


Repository Staff Only: item control page