Hentze, Konrad Friedemann: Suitability Analysis of Satellite Remote Sensing Methods to Map Agricultural Land Use Change after Zimbabwe's "Fast Track Land Reform Programme". - Bonn, 2017. - Dissertation, Rheinische Friedrich-Wilhelms-Universität Bonn.
Online-Ausgabe in bonndoc: https://nbn-resolving.org/urn:nbn:de:hbz:5n-46988
@phdthesis{handle:20.500.11811/7173,
urn: https://nbn-resolving.org/urn:nbn:de:hbz:5n-46988,
author = {{Konrad Friedemann Hentze}},
title = {Suitability Analysis of Satellite Remote Sensing Methods to Map Agricultural Land Use Change after Zimbabwe's "Fast Track Land Reform Programme"},
school = {Rheinische Friedrich-Wilhelms-Universität Bonn},
year = 2017,
month = apr,

note = {Forced evictions of white commercial farmers in the year 2000 took the Zimbabwean land reform programme to a new level. While international media covered the tragedies of white farmers, the collapse of a state economy and the food shortages in the country, new farmers emerged on former large scale production schemes. At the same time, the Zimbabwean Government officialized the community driven farm take overs and provided legislation and administration through a 'Fast Track Land Reform Programme' (FTLRP). In recent years, there has been a vivid scientific debate on whether new farmers are productive and on whether the FTLRP has been successful or not. But despite the remarkable reception in media and academic work, actual figures are hard to get hold of.
Aiming to provide answers to the often posed question of 'what has happened where after the FTLRP?', the present thesis provides an analysis of geospatial methods to research tenure change and its consequences. It conceptualizes the potential of remote sensing and GIS to contribute to an integrated land reform research, and assesses whether this potential has been fully tapped. An analysis framework for the FTLRP is presented to overcome the apparent lack of spatial research. In order to proof its applicability, three of the proposed spatiotemporal methods are critically examined for their suitability to adequately map outcomes of the land reform: phenology based temporal land use classification, productivity trend mapping, and breakpoint analysis.
A time-series based multi-temporal land use classification approach on Moderate Resolution Imaging Spectroradiometer (MODIS) data was used to analyse trajectories of the agricultural sector. The objective of this approach was to produce a dataset of annual binary crop/non crop maps for former freehold tenure areas, providing spatial and temporal information on Zimbabwe's agricultural production after the FTLRP. However, the well-established and commonly used method of phenology based land use classification does not provide meaningful results. Therefore, after a comprehensive review of limitations of MODIS-NDVI based hard classifications, a relative assessment approach is additionally presented. As answers to questions of Zimbabwe's national agricultural production after the FTLRP remained unanswered, trend analysis of vegetation was applied for crop area of Zimbabwe's most productive farming region. The results reveal that no correlation between negative production trends and new land tenure types exists. Trends of production are rather characterised by spatial clusters of strong negative trends resulting from operation stops of irrigation schemes.
To map fallow irrigation schemes in Zimbabwe, a hybrid-methodology of “Breaks For Additive Seasonal and Trend” (BFAST) and “Seasonal Trend Analysis” (STA) was introduced. The combination of 'time of peak' calculated with STA, and 'breakpoint of seasonal signal' calculated with BFAST, provides sound results for time and location of abandonment of irrigation schemes which had great impacts on the country's food and export crop production. It is concluded that temporal classification results based on medium scale resolution data have to be used with great caution and that combination of methods significantly enhances explanatory power of MODIS-NDVI data. By using FTLRP as a case study for the suitability of remote sensing to map tenure induced land use changes, the thesis forms an essential contribution towards integrated land use science.},

url = {https://hdl.handle.net/20.500.11811/7173}
}

The following license files are associated with this item:

InCopyright