CD Skripsi
Analisis Prediksi Kedalaman Muka Air Tanah Di Lahan Gambut Menggunakan Data Historis (Studi Kasus Di Khg Pulau Bengkalis)
ABSTRACT
Almost every year during the dry season, forest fires in peatlands occur, with the
main factor being drought. In peatlands that have undergone land use change, there
is excessive construction of drainage canal networks that accelerate the flow of
water out of the peat. Depth to Groundwater Table is a key parameter in
overcoming these problems. The aim of this research is to create a hydrological
model using regression analysis that is used to predict the water table in peatlands,
so as to predict the level of fire risk. The data used for this modeling are historical
data of groundwater level fluctuations and rainfall recorded from SESAME at 4
Bengkalis Island KHG stations, as well as supporting data on GPM satellite
rainfall. Regression analysis was conducted using four data length scenarios,
namely three months, six months, nine months, and twelve months to determine
which time period can represent hydrological conditions in the field. The results
showed that the regression analysis equation using nine months of data length at
Pendekik and Penampi Stations with a correlation coefficient of 0,9620 and 0,9669,
six months of data length at Bantan Tua Station with a correlation coefficient of
0,8608, and three months of data length at Muntai Station with a correlation
coefficient of 0,9443. From the regression analysis equation obtained for each
station, Pendekik Station shows the highest drainage rate when there is no rain in
a two-week period of 8,87 cm/day. In this study it can be concluded that the
regression method with a conceptual model is able to make modeling in the
prediction of groundwater level elevation for the future with a certain time range.
Keywords: Correlation value, modeling, regression method, water table.
Tidak tersedia versi lain