Integrated spatial layers for delineating grazing units in central Borana zone of southern Ethiopia
Description
Database derived by summing up three information layers, namely topographic features, vegetation cover and its dry-out dynamics. Three number codes were adopted to sum up values on a pixel basis and the place value of the number in the code is used to interpret individual pixel behavior. With four categories of topographic ruggedness index, four categories of vegetation greenness index and five categories of vegetation dry-out dynamics, we came up with a thematic layer with 80 classes. In this database, spatial layer is oriented towards topographic features. Resource users can combine these thematic classes to come up with the number of rangelands management units that suits their purposes.
Files
Steps to reproduce
Each of the three spatial layers is separately derived from their respective data sources. Topographic Ruggedness Index (TRI) was created from SRTM digital elevation model. Algorithm for creating TRI was published by Riley et al. 2019. (see for example Riley, S., D. Stephen & E. Robert (1999). A terrain Ruggedness Index That Quantifies Topographic Heterogeneity. Intermountain Journal of Sciences, 5, 23-27. TRI values were classified into four categories using quantile thresholding in ArcMap 10.2 and each class was assigned a score starting with 100 up to 400 in a step of 100. Vegetation greenness index was integrated in the dataset and this was calculated using perpendicular vegetation index that minimizes background reflectance effect. PVI was calculated using the algorithm published by Richardson and Wiegand 1977 ( see for example Richardson, A. J. & C. L. Wiegand (1977). Distinguishing vegetation from soil background information (by gray mapping of Landsat MSS data). Photogramm. Eng. Remote Sens., 43, 1541-1552). Resulting layer was classified into four vegetation greenness categories using one standard deviation thresholding technique in ArcMap 10.2 with codes ranging from 10-40 with 10 indicating sparse vegetation and 40 dense vegetation. Vegetation dry-out dynamic was also separately calculated on stack images by fitting linear regression model and resulting layers classified using quantile thresholding into five vegetation dry-out dynamics and then each layer was assigned a code ranging from 1-5 with 1 indicating slow out dynamics and 5 indicating fast dry-out nature. The three information layers are summed up to come up with a three digit code that is easy to interpret. For example code 145 is interpreted as a nearly flat ground, dense vegetation with fast dry-out dynamics. With this kind of satellite data interpretation scheme, resource users are able to interpret individual pixels with minimal ground data and only local knowledge is required.