UJI PERFORMA CITRA TIME-SERIES COMPOSITE CLEAN-COASTAL-WATER SENTINEL-2 UNTUK PEMETAAN BATIMETRI DI PERAIRAN DANGKAL

Munawaroh, Prof. Dr. Pramaditya Wicaksono, M.Sc; Dr. Nur Mohammad Farda, M.Cs

2024 | Tesis | S2 Penginderaan Jauh

Algoritma pra-pemrosesan citra “time-series composite clean-coastal-water” merupakan salah satu alternatif untuk memperoleh citra satelit yang minim gangguan atmosfer untuk input data pemetaan batimetri perairan dangkal dari citra satelit. Akan tetapi, performa dari citra time-series composite untuk pemetaan batimetri perairan dangkal perlu diuji, mengingat adanya disperse piksel dan band pada saat proses pembangunan citra komposit. Tujuan dari penelitian ini adalah adalah mengkaji hasil citra time-series composite clean-coastal-water Sentinel-2 berbasis cloud computing sebagai data input untuk pemetaan batimetri di perairan dangkal, membandingkan hasil ekstraksi batimetri perairan dangkal dari citra Sentinel-2 perekaman tunggal dan citra time-series composite clean-coastal-water menggunakan algoritma band ratio model dan random forest regressor dari segi akurasi dan distribusi spasial kedalaman perairan dangkal. Dan mengevaluasi performa dan kualitas peta batimetri perairan dangkal yang dihasilkan dari citra perekaman tunggal dan citra time-series composite clean-coastal-water Sentinel-2 berdasarkan standar International Hydrographic Organization (IHO). Hasil penelitian menunjukkan bahwa tidak semua proses time-series composite dapat menghasilkan citra yang bersih dan bebas dari gangguan awan, sunglint, perairan keruh, dan piksel pecah gelombang. Rentang waktu perekaman berpengaruh terhadap kualitas citra time-series composite yang dihasilkan. Model SDB RFR dapat menangani hubungan kompleks antara variable kedalaman perairan in situ dengan nilai surface reflectance dari citra time-series composite clean-coastal-water yang memiliki ketidakpastian dispersi piksel dan band daripada model SDB BRM. Secara keseluruhan, kualitas data SDB yang dihasilkan dari metode BRM belum memenuhi syarat ketelitian data kedalaman perairan dangkal menurut CATZOC IHO. Model SDB RFR berhasil digunakan dalam memprediksi batimetri di perairan dangkal dengan akurasi yang memenuhi syarat CATZOC standar IHO, yaitu nilai RMSE dan MAE kurang dari 0,5 meter juga nilai tingkat kepercayaan TVU lebih dari 95% pada kedalaman 0-5 meter yang dihasilkan dari masing-masing citra time-series composite CCW di Tanjung Kelayang, Kepulauan Seribu, perairan sekitar Pulau Morotai dan perairan sekitar Pulau Ontoloe.

The "time-series composite clean-coastal-water (CCW)" image pre-processing algorithm is one of the alternatives for obtaining minimal atmospheric disturbance satellite images for data inputs in satellite Derived Bathymetry (SDB). However, the performance of time-series composite images for shallow water bathymetric mapping needs to be assessed, considering the presence of pixel and band dispersion during the composite image construction process. This research aims to study the result of a time-series composite clean-coastal-water Sentinel-2 by using cloud computing pre-processing as input data for SDB in shallow waters, compare the results of SDB from a Sentinel-2 single date image and from the time-series composite clean-coastal-water image using a band ratio model (BRM) and a random forest regressor (RFR) algorithm in terms of accuracy and spatial distribution of the bathymetry, and evaluate the performance and quality of bathymetric maps obtained from a single date image and time-series composite clean-coastal-water Sentinel-2 images based on International Hydrographic Organization standards (IHO). The results show that not all time-series composite processes can produce images that clean of interference from clouds, sunglints, bad waters, and pixel wave breaks. The acquisition time range affects the quality of the time-series composite image. The SDB RFR model can handle the complex relationship between the in situ water depth variable and the surface reflectance value of a time-series composite clean-coastal-water image that has pixel and band dispersion compared to the SDB BRM. Overall, the quality of the SDB data generated by the BRM method has not qualified for the scale of data on shallow water bathymetry according to CATZOC IHO standards. The SDB RFR model was successfully used to predict bathymetry in shallow waters with accuracy that meets the IHO standard CATZOC requirements, with RMSE and MAE values of less than 0.5 meters as well as TVU confidence level values of more than 95% at a depth of 0-5 meters resulting from each - respectively CCW composite time-series images in Tanjung Kelayang, Seribu Islands, waters around Morotai Island and waters around Ontoloe Island.

Kata Kunci : Satellite Derived Bathymetry, shallow water, Sentinel-2, clean-coastal-water, time-series composite