Analisis Distribusi Spasial dan Tren Konsentrasi Total Suspended Solid (TSS) Menggunakan Citra Sentinel-2 dan Data In Situ

Abstract

Sungai Cisadane di kawasan Pintu Air 10 merupakan salah satu sumber air baku utama bagi masyarakat Tangerang, namun aktivitas masyarakat menyebabkan penurunan kualitas air. Total Suspended Solid (TSS) merupakan parameter penting yang mencerminkan konsentrasi padatan tersuspensi di perairan. Penelitian ini bertujuan mengevaluasi akurasi estimasi TSS menggunakan citra Sentinel-2 terhadap data in situ, menganalisis distribusi spasial TSS, serta mengkaji tren perubahan TSS periode 2017–2026. Data yang digunakan meliputi 30 sampel in situ dengan metode gravimetri dan citra Sentinel-2 MSI. Estimasi TSS dilakukan menggunakan algoritma Liu, Nechad, Budiman, Parwati, dan algoritma lokal berbasis regresi linear. Evaluasi dilakukan menggunakan koefisien determinasi (R²), Root Mean Square Error (RMSE), dan Mean Absolute Error (MAE). Hasil penelitian menunjukkan bahwa algoritma lokal Cisadane, yaitu TSS = 82,43 × (B5/B2) - 50,13, memberikan performa terbaik dengan R² = 0,449, RMSE = 13,75 mg/L, dan MAE = 12,46 mg/L. Distribusi TSS didominasi kelas rendah (25–50 mg/L), sedangkan kelas sedang (50–100 mg/L) terutama ditemukan di hilir Pintu Air 10. Tren TSS selama 2017–2026 berfluktuasi dengan konsentrasi tertinggi pada tahun 2019. Kata kunci: Sentinel-2, Sungai Cisadane, Total Suspended Solid, Penginderaan jauh, Algoritma lokal

The Cisadane River in the area of Sluice Gate 10 is one of the main sources of raw water for the people of Tangerang, but community activities have caused a decline in water quality. Total Suspended Solid (TSS) is an important parameter that reflects the concentration of suspended solids in waters. This study aims to evaluate the accuracy of TSS estimation using Sentinel-2 imagery against in situ data, analyze the spatial distribution of TSS, and examine the trend of TSS changes for the period 2017–2026. The data used include 30 in situ samples with the gravimetric method and Sentinel-2 MSI imagery. TSS estimation was carried out using the Liu, Nechad, Budiman, Parwati algorithm, and a local algorithm based on linear regression. Evaluation was carried out using the coefficient of determination (R²), Root Mean Square Error (RMSE), and Mean Absolute Error (MAE). The results showed that the local Cisadane algorithm, namely TSS = 82.43 × (B5/B2) - 50.13, provided the best performance with R² = 0.449, RMSE = 13.75 mg/L, and MAE = 12.46 mg/L. The TSS distribution was dominated by low class (25–50 mg/L), while medium class (50–100 mg/L) was mainly found downstream of Water Gate 10. The TSS trend during 2017–2026 fluctuated with the highest concentration in 2019. Keywords: Cisadane River, Local algorithm, Remote sensing, Sentinel-2, Total Suspended Solids