Integrating crop modelling with remote sensing : Application on data-limited small-scale farming systems
Summary
No summary yet. VIRAT can read this paper and write a plain-language summary of its aim, methods, key findings and limitations.
Summarise with VIRAT Free to try with your Science Catalogs account.Abstract
Climate change and variability present increasing threats to agricultural productivity and food security in southern Africa, particularly for small-scale, rainfed farming systems constrained by limited resources, fragmented landholdings, and weak institutional support. Although digital technologies such as crop growth models (CGM) and remote sensing (RS) show growing potential for improving yield estimation and adaptation planning, their practical use in small-scale contexts remains limited. This thesis examined how CGM and RS can be integrated and applied in data- and resource-limited small-scale maize farming systems to improve yield estimation, climate adaptation, and institutional decision-making. The study was conducted in South Africa’s Eastern Cape Province, a region that exemplifies the socio-economic and climatic challenges of small-scale agriculture in southern Africa. A mixed-methods approach was used, combining a systematic review, crop modelling experiments, climate impact assessments, and institutional surveys. Chapter 2 provided a systematic overview of global CGM–RS integration through data assimilation, identifying key trends, limitations, and research gaps, particularly the concentration of applications in large-scale, high-input systems and the lack of studies in African small-scale contexts. Building on these insights, Chapter 3 applied a recalibration-based data assimilation method using Sentinel-2–derived Leaf Area Index (LAI) to improve the performance of the WOFOST crop model at the farm scale. The model was parameterized and validated using field data from small-scale maize plots, demonstrating that assimilating LAI significantly improved yield estimation accuracy by more than 90% compared to baseline simulations. Chapter 4 used this locally recalibrated model to simulate maize phenology, water stress, and yield responses across three future climate scenarios derived from multiple Global Climate Models (GCMs) and Shared Socioeconomic Pathways (SSPs). Results indicated that projected increases in temperature and rainfall variability will substantially reduce maize yields, though adaptive measures such as adjusted planting dates and changing varieties can partially offset these losses. Chapter 5 assessed the awareness, use, and institutional readiness of agricultural extension officers to adopt CGM and RS tools. Survey and interview data revealed high interest but low practical use, primarily due to limited training, technical complexity, and weak institutional infrastructure. Overall, the findings show that integrating CGM with RS is technically feasible and operationally valuable in small-scale systems, provided it is supported by institutional capacity-building and policy alignment. Methodologically, introducing a yield gap factor and assimilating high-resolution RS data are pragmatic innovations for adapting modelling frameworks to data-poor environments. Scientifically, the research advances understanding of how digital tools can improve yield forecasting and climate adaptation at local scales. Practically, it highlights the pivotal role of extension officers in bridging the gap between digital innovation and on-farm application. By linking the technical, climatic, and institutional dimensions, this thesis offers a pathway to democratize digital agriculture and build more resilient, data-informed small-scale farming systems across southern Africa.
Fields of Study & Subjects
Remote Sensing in Agriculture
Climate change impacts on agriculture
Smart Agriculture and AI
Publication details
- Publication date
- 1 January 2026
- Language
- EN
- Document type
- Thesis