Integration of seasonal forecast information and crop models to enhance decision making in small-scale farming systems of South Africa
Thesis
Permanent link to this Item
Authors
Supervisors
Journal Title
Link to Journal
Journal ISSN
Volume Title
Publisher
Publisher
Faculty
License
Series
Abstract
Climate variability threatens agricultural productivity and household food security, amongst small-scale farmers of South Africa. Managing climate variability is challenging due to the variation of climate parameters and the difficulty in making decisions under such conditions. Integrated seasonal forecast information and crop models have been used as a tool that enhances decision making in some countries. Utilization of such an approach in South Africa can enhance decision making in climate variability management. The study therefore sought to formulate a decision-making approach to enhance climate variability management in small-scale farming systems of South Africa through integrating seasonal forecast information and crop models. Current practices, challenges and opportunities for climate variability management by different small-scale farmer types were identified using focus group discussions and local agricultural extension officers. The Climate Forecast System version 2 (CFSv2) model-based forecasts were integrated with the Decision Support System for Agrotechnology Transfer (DSSAT) v4.7, a mechanistic crop model based on the Global Climate Model (GCM) approach. The GCM approach was the most appropriate technique for integrating seasonal forecast information and the crop model due to the compatibility in the forecast and crop model format. The decision-making process was formulated through assessing the simulation yield patterns under a range of farm management practices and seasonal forecasts for different cropping seasons, crops and farmer types for Limpopo and Eastern Cape, South Africa for 2017/18 season. The study assessed 48 different potential combinations of farm management practices: organic amendments, varieties, fertilizers and irrigation. Benefits of the decision formulation process and specific seasonal forecast-based recommendations were then assessed in the context of the performance of the practices under historical measured data for the conditions; 2011-2017, using percentile ranking. Assessing the yield response patterns under different farm management practices and seasonal forecasts (2017/2018), the study realized a range of decision scenarios. These are (1) low decision capacity and low climate sensitivity where there is low value for decision due to the homogeneous performance of the different management practices given climate forecasts. (2) high decision capacity and low climate sensitivity, where there is higher potential value for decision making as the different practices have uniform performance across climate forecasts. (3) High decision capacity and high climate sensitivity, where the good response to change in practices under changing climate forecasts. Confidence in the decision formulation process v was re-enforced as some of the decision scenarios were also realized under different conditions in the period; 2011-17. The scenario (2): High decision capacity and low climate sensitivity was predominant in locations with low forecast skill. In contrast the scenario (3): High decision capacity and high climate sensitivity was predominant in locations with high forecast skill. The decision formulation process allows for assessment of farm management practices in the seasonal forecast decision space. Although the case study realized some scenarios ahead of others, the process is robust and repeatable under any conditions. Although the process does not always offer recommendation with improved value for decision making, the value of recommendations is greater under decision scenarios with greater decision capacity. Such benefits are crop and location dependent. Improved seasonal forecasting skill increases reliability of the decision-making process, decision scenarios and associated recommendations. Such assertions need to be tested on the field scale to assess their practical feasibility.