Evapotranspiration (ET) is one of the climate elements which plays an important role in ecosystem water balance, including in oil palm plantations. Therefore, many mathematical equations and algorithms have been developed and designed to estimate and determine the spatial distribution of evapotranspiration. Remote sensing data are one of the important sources and techniques to estimate spatial variation of various climate elements, including ET. The main objective of this research is to estimate the spatial variation of ET using the SEBAL algorithm and Landsat-8 imagery of a large-scale commercial oil palm plantation, i.e. PT Perkebunan Nusantara VI Batanghari, Jambi Province, Indonesia. The analysis is carried out using Landsat-8 (OLI/TIRS) data and reference meteorological data from a micrometeorology flux tower. We calculated surface radiance, reflectance, albedo, Normalized Difference Vegetation Index (NDVI), emissivity, surface temperature, net radiation, soil heat flux, sensible heat flux, and latent heat flux to derive the hourly and daily evapotranspiration from the study area. Validation of ET from the SEBAL model were performed against ET from aerodynamic measurements from a micrometeorological tower at the same site. Differences in ET within “only oil palm cover” are relatively low and that difference in ET over the entire area of the oil palm plantation is mainly between oil palm vs. open lands, roads, and buildings. The evapotranspiration values of oil palm cover (NDVI 0.45-0.54) were between 2.42 ± 0.36 – 3.36 ± 0.17 mm d-1 . There was no significant difference between ET derived from SEBAL compared to aerodynamic methods (p-value = 0.598; r = 0.75).
Changes in land use and land cover play a critical role, especially in the Lore Lindu National Park (TNLL) area, impacting on ecosystem functions. This study was aimed at analyzing the dynamics of land change and the factors influencing the land change in this National Park. The methods used were the GIS technique and a binary logistic regression model. The land changes locating between Sigi Regency and Donggala Regency, Central Sulawesi Province, Indonesia, which consisted of thirteen sub-districts in the TNLL region acquiring from Landsat satellite data acquisition of 1997, 2002, 2013, and 2018. The dynamics of land changes during the period of 1997-2002 has decreased the forest area by 2,643 ha, and the next period, 2002-2013, the decline of forest area has reached 4,265 ha, and the overall the dynamics of land change experiences a significant increase from 1997 to 2018 which declines the forest areas approximately 10,175 ha and followed by a decline of the meadow area in 1,726 ha changing function into the built-up land of 526 ha, mixed gardens of 1,189 ha, fields/moorings of 3,019 ha, rice fields of 1,548 ha and shrubs of 5,619 ha. The factors influencing the land change in this TNLL region based on the results of binary logistic regression analysis are the population density (X5), distance from the settlement (X3), distance from the road (X2), distance from the capital (X4), and topographic conditions (X1). Of these five variables, the population density has the highest negative regression coefficient, which is equal to -0.068. The regression equation is Y = -0.094X1-0.157X2-0.176X3-0.083X4- 0.068X5 and being significant in a level of 0.001 percent that indicates these five factors have influenced greatly high the land change in the TNLL. This situation can be inferred that the free distribution and population growth in the National Park influence increasing the conversion of forest areas.