Integrating support vector machine and cellular automata for modelling land cover change in the tropical rainforest under equatorial climate in Ghana

Abstract

Unsustainable anthropogenic activities such as indiscriminate logging of trees, mineral exploitation, conversion of forest into agricultural lands are known to cause major environmental changes, thereby triggering a chain of irreversible forest depletion. This has called an urgent need by government and private agencies to institute policies and programs to curtail the destruction of the ecosystem due to the pressure on the available land. In this study, the Land use/land cover changes between the period of 1986 and 2020 in the tropical rainforest of Ghana was considered. A combination of machine learning and Markov chain approach was adopted to project future LULC for 2040 and 2060.The results showed that area covered by Open Forest declined from 21,531.87 km2 to 14,518.82 km2 and Dense Forest also declined from 14,313 km2 to 8202.98 km2 over a period of 34 years. The CA-Markov model was used to predict the future land use land cover, and it was observed that the total forest cover could decline to 15,551.79 km2 in 2040 and further decrease to 13,401.79 km2 in 2060. It was also found that settlement, mining and agricultural land, which is be driven by rapid population increase, has contributed significantly to the rapid declining forest cover. The results of this study have demonstrated the impact of unsustainable use of natural resources in these three regions. It also highlights the need for concerted effort to develop comprehensive environmental policies to encapsulate sustainable conversion and utilisation of natural resources by focusing on water-energy-food nexus.

Keywords

• Support vector machine
• Tropical rainforest
• CA-Markov
• Land use change