Crop change detection in conflict areas, Nigeria, October 2020 (2020-10-19)

Activation date: 2020-10-19
Event type: Humanitarian

Activation reason:

The scope of the activation is to support the WFP officers with geospatial analysis regarding the agriculture and food security situation and nutrition analysis in Borno state in north-eastern Nigeria. Due to the inaccessibility of the region caused by on-going armed conflicts, the impact of the conflict needs to be estimated for changes in cropland (loss/gain) and affected population in case of cropland loss. The current activation is an update of EMSN-063, assessing changes between 2020 & 2019 and 2020 & 2010.The process steps followed to implement the analysis are:Generation of NDVI-composite layers per AOI for 2020. Change detection of crop areas was performed using NDVI composite raster layer based on HR imagery data. As single NDVIs provide an indication of the presence of vegetation and its photosynthetic activity, it is not sufficient to distinguish crops from surrounding natural vegetation. Compared to this natural vegetation, cropland is identified by its high changes of NDVI values between the different phases of the vegetation period: Photosynthetic activity of crops are low during the ploughing and sowing period, increase during the growing phase until reaching a maximum value right before the harvest. Once harvested, NDVI values should decrease drastically. ​Update and verification of the populated sites layer based on freely available imagery and data. To assess the possible impact of the population affected inside the AOIs, several credited and accurate population datasets that depict the distribution of the population on a global scale, were analysed. WorldPop dataset was the source selected to provide population counts per site. This dataset represent estimates of population counts in wall-to-wall coverage at a 100 m spatial resolution. To allocate the spatially distributed population estimates to each of the population sites, that are spatially represented as a point feature it was necessary to interpolate an area of influence for each one of the 2,616 sites identified in the 5 LGAs of analysis.Interpolation of the population data per populates site applying the Thiessen polygon analysis. Thiessen polygons analysis was performed to determinate the allocations areas per each population site. This method generated several partitions of an area ("catchments") where any location within these smaller regions is closer to its associated input point than to any other point. Therefore, for each site it was generated an allocation area making possible to allocate the population counts inside it to its corresponding population site.Crop change analysis per populated site based on the NDVI-composite layer between 2020 & 2019 and 2020 & 2010.  Analysis of population affected by cropland loss. Comparing the changes from 2010-2020 with 2019-2020, it is clear that more than two thirds of cropland loss occurred before 2019, as more than 70% of the locations considered show no change between 2019 and 2020. Of these, 75% were sites affected by cropland loss over the period 2010-2020, 18% of the sites show no change and only 7% of the sites show cropland increase over the period 2010-2020. It can therefore be assumed that the majority of the sites analysed are still affected by the effects of the cropland loss occurring up to 2019.​

European Commission, Joint Research Centre (JRC) (2020): Crop change detection in conflict areas, Nigeria, October 2020 (2020-10-19). European Commission, Joint Research Centre (JRC) [Dataset] PID: http://data.europa.eu/89h/d3550b3d-3499-43d2-8a96-5cd09fa19b28

No additional information