Simulating Agricultural Land-Use Adaptation Under Changing Climate Using Multi-Agent System Model in the Upper East Region of Ghana

Abstract

One of the operationalised tools of human-environment system (HES) modelling approach is Multi-Agent System (MAS) which has been used in a number of areas to study the dynamics and management of human and natural systems especially when facing unexpected disturbances. Therefore, increasing studies are interested in using Multi-Agent analyses for the understanding of agricultural adaptation to environmental changes.

However, when it comes to the use of MAS for the operationalisation of adaptation decision making in agricultural land use based on farmers’ perception of climate variability, only very few studies empirically operationalise the concept in their simulations. Also, anotherchallenge is how to isolate planned adaptation within a large traditional number of autonomous adaptation practices. The current research therefore focused on the implementation of a MAS approach for investigating the traditional adaptive strategies in a small scale area in the Upper East Region of Ghana by considering farmers’ perception of climate change and variability. In order to achieve the purpose of this, Land Use Dynamic Simulator (LUDAS) approach was adapted and modified by integrating the two stepdecision making sub-models. This modified version of LUDAS called SKY-LUDAS (referring to the communities where it was implemented: Sirigu-Sumbrungu-Kandiga-Yuwa) was constructed to capture the empirical heterogeneity of farm household agents and landscape agents (biophysical environment), and also to explicitly simulate interactions between these two agent types.

From the results of the multivariate statistical methods, three farm household agent groups were identified. Also the factors explaining the decision of these three household agent groups on the choice of the six identified land-use types were analysed. Two submodels were developed and calibrated for implementing the two-step decision making submodels:

Perception-of-Climate-Change and Adaptation Choice strategies. Simulation results of SKY-LUDAS suggested that the land-use behaviour in the study area reflects a tendency of subsistence farming. In terms of farm-households’ livelihood strategy, especially the structure of the gross income, there was a growing contribution of rice and groundnut. Also the pattern of the gross income under the scenario of perception on climate change (PCC) showed explicitly the contribution of the adaptation options in the households’ livelihood strategy. Accordingly, SKY-LUDAS has revealed a gradual shift among land-use types from traditional cereals farming to the cultivation of groundnuts, rice, maize and soybean.