Abstract

Intercomparison of spatial input datasets to crop models and implications
Track: Sustainable Development and Humanitarian Affairs
Authors: Jawoo Koo

Crop models simulate growth and yield of crops based on the field-observed input datasets, such as soil, weather, plant, and farmer's field management practices. The result of simulation can be interpreted as the potential crop productivity changes under different scenarios of farm management and environmental impacts, and used to advise farmers and make an informed decision on farm management. When the models are used at grid-based regional scale, soil and weather data are not measurable in the field; they are often retrieved from global datasets available at different formats and coarser resolutions. Using a sub-Saharan Africa-wide crop systems modeling framework developed by the HarvestChoice project, this paper discusses the implication of using a particular soil and climate dataset, versus others, on the final outcome and its potential influence in the interpretation and investment decisions, and suggest the best practices to avoid introducing biases in the analysis.