Pemanfaatan Penginderaan Jauh Untuk Pemantauan Penurunan Muka Tanah Dan Korelasi Terhadap Perubahan Lahan Di Kota Pekalongan Tahun 2016-2022
Asy Syifa Sekarariani, Dr. Ir. Harintaka, S.T., MT., IPU. ASEAN Eng.,
2023 | Skripsi | TEKNIK GEODESI
Pekalongan City is one of the strategic areas on the north coast of Java Island which is exposed to the danger of land subsidence. Population density tends to increase with the increase in population. From 2016 to 2022 there will be significant growth reaching 310,869 thousand people. This affects the mass of the building which is getting bigger so that it causes natural compression on this alluvial plain. The impact of this phenomenon contributes to each other to influence spatial management and development. Population growth and infrastructure cause land cover so monitoring needs to be done to obtain information on land changes. Therefore, this study aims to examine the relationship between land subsidence and changes in land use in 2016, 2019, and 2022.
In this study, 3 pairs of SAR Sentinel-1 satellite imagery data were used to calculate land subsidence using the DInSAR method in SNAP software. The DEM data used is DEMNAS 0,27 -arcsecond area with a resolution of 8 meters. To identify changes in land cover using the Sentinel-2 dataset available by Google Earth Engine with guided classification and random forest algorithms. Class classification is divided into 4 classes, namely open land, water bodies, vegetated areas, and settlements. These two results were visualized and statistically calculated with QGIS software to see the correlation of land subsidence results and land cover.
The results of this study for PMT values for the 2016-2022 range yield values of 0 to 22 cm/year with a significant decrease in values in 2019 with an average speed of 8,18 ± 4,68 cm/year. The average speeds for 2016 and 2022 are respectively 1,54 ± 1,93 cm/year and 2,66 ± 1,91 cm/year. The largest average decrease was in West Pekalongan Sub-District of 12,75 ± 2,82 cm/year. The smallest average decrease was in East Pekalongan District of 0,45 ± 0,76 cm/year. Statistical tests of the DInSAR processing data and GNSS data yield an RMSE value of 6,44 cm with a standard deviation of 4,42 cm. The results of land cover classification with significant changes in area were in 2019 in the class of settlements or built-up areas of 50,54% of the total area of the study area. The land cover accuracy test was carried out by field checking (ground truth) which resulted in an overall accuracy value of 87,88% and a kappa accuracy of 0,84. The relationship between PMT and the area of change in land cover shows that the highest subsidence is dominated by the settlement class which has an impact of 2349 ha with the highest land subsidence reaching a value of 9,26 cm/year. The two variables have a positive correlation.
Kata Kunci : Penurunan Muka Tanah, Sentinel-1, DInSAR, SNAP, Tutupan Lahan, Sentinel-2, Random Forest, Google Earth Engine