Prediksi Kedalaman Muka Air Tanah Di Lahan Gambut Menggunakan Metode Regresi Di Khg Pulau Tebing Tinggi

Abstract

The forest fires that occurred in 2015 were very severe in the past 20 years. The forest fires caused a haze disaster for several months. The depth of groundwater plays an important role as a key in determining the risk of a fire occurring on peatlands. The purpose of this study is to create a hydrological model to predict the depth of the groundwater surface as a parameter for potential fires on peatlands using regression analysis. The location of the research is the village of Sungai Tohor. The data used for modeling are rainfall data and groundwater depth data that have been recorded historically on peatlands by installing SESAME at the research location and obtained data through the website address web.sesame-system.com. Regression analysis was performed using five data length scenarios, namely 1 month, 3 months, 6 months, 9 months, and 12 months to determine the period that could represent the hydrological conditions in the field. The results showed that the regression analysis with data lengths of 1 month, 3 months, 6 months, and 12 months gave the best prediction results for groundwater levels with strong correlation coefficient values of 0.80, 0.93, 0.89 and 0.94. Meanwhile, groundwater level prediction using regression analysis with 6 months length of data gives results with a moderate correlation value of 0.56.
Key words: correlation value, groundwater level, peatlands, regression analysis.