Abstract:
Tropical deforestation or degradation is on the increase majorly due to anthropogenic factors such as urbanization, illegal felling, agricultural expansion, fires, animal grazing and infrastructural expansion. These changes are often difficult to monitor except with the use of satellite remote sensing. The difficulties associated with assessing satellite images in developing countries led to the introduction of the Web-based Earth Observation analytical tools called SEPAL.IO by the Food and Agriculture Organization of the United Nations. In this study, SEPAL.IO toolset was used to analyze change detection in Oyinmo Forest Reserve over a period of thirty-two years (1986 – 2018). The main objective of this research work is to provide a reliable dataset to support the development of early warning system in Oyinmo Reserve. Land use / Land cover analysis was carried out through supervised classification. Change detection analysis classified the forest area into four (4) different classes of Stable forest, Stable Non-Forest, Forest loss and Forest gain. The result further showed that Stable Forest occupied an area of 2.50 km2, Stable Non-Forest 5.80 km2, Forest Loss 16.17 km2 and Forest Gain 9.06 km2 for the Epoch 1 (1986 – 2002) and Stable Forest 7.20 km2, Stable Non- Forest 9.31 km2, Forest Loss 11.00 km2 and Forest Gain 6.01 km2 for the Epoch 2 (2002 –2018). Findings showed an annual growth and loss rates of 3.01% and 1.69 % respectively for Epoch 1 and 2.05 % and 1.12 % respectively for Epoch 2 recorded. This study also proved the usability and applicability of SEPAL.IO remote sensing toolset for detecting forest change over a certain period of years. An overall accuracy for the LULC classification was recorded as 77.5%, 82.5% and 80.0% for 1986, 2002 and 2018 respectively and a Kappa Coefficients of agreements ranged from 0.55 to 0.65.
