Jurnal Teknik Sumber Daya Air

Penerapan Deep Learning untuk Prediksi Tinggi Muka Air Sungai dengan Mempertimbangkan Faktor Operasi Bendungan

Mamad Tamamadin — 2025-12-31

Indonesia is a flood-prone region, making early detection through accurate river water level prediction essential by utilizing efficient modeling methods. This study aims to develop a high-accuracy water level (TMA) prediction model at several river flow monitoring stations by employing spatial rainfall map datasets and dam discharge information as input variables in a deep learning framework that combines CNN and LSTM architectures. The model was tested under two scenarios, with and without dam operation, and its predictive performance was evaluated at three monitoring sites (Katulampa, Kampung Kalapa, and MT. Haryono). The initial evaluation using only spatial rainfall input for the MT. Haryono station showed a correlation coefficient of 0.65, MAE of 0.412 m, and NSE of 0.58. After incorporating multiple rainfall images and integrating discharge data based on dam operation scenarios, a significant improvement in prediction accuracy was observed across all stations, with the correlation increasing to 0.88, MAE decreasing to 0.137 m, and NSE rising to 0.85. These findings confirm that the inclusion of additional hydrological information, particularly dam operation data, can substantially enhance the reliability of river water level prediction models.

Analisis Komparatif dan Validasi Metodologi Delineasi DAS Otomatis Berbasis Artificial Intelligence (AI)

Lutfi — 2025-12-31

Accurate watershed delineation is a fundamental stage in hydrological modeling. However, conventional methods based on the Deterministic 8 (D8) algorithm often suffer from reduced accuracy in complex landscapes and areas affected by human activities. This study presents a comparative performance analysis of three delineation approaches, all of which utilize the D8 algorithm with high-resolution DEMNAS as input. The approaches include the standard implementations in HEC-HMS and WMS software, as well as an implementation on a newly developed innovative platform, ACAP (Advanced Catchment Analysis Platform). ACAP leverages Artificial Intelligence (AI) to pre-process and refine DEM data by integrating information from satellite imagery. These three approaches were tested on three dam catchment areas in Indonesia with diverse characteristics: Napun Gete, Batutegi, and Greneng. Their spatial accuracy was subsequently evaluated using the Jaccard Index (IoU) metric against ground truth data. The findings indicate that while HEC-HMS and WMS are highly reliable in natural watersheds, the ACAP platform demonstrated significant superiority in the complex Greneng catchment, achieving the highest IoU score (0.98) compared to the conventional platforms (0.84-0.85). This study concludes that the most impactful innovation for enhancing D8 delineation accuracy is not replacing the algorithm itself, but rather empowering it through smarter input data preparation. This highlights the significant potential of AI-augmented approaches for future precision hydrological analysis.

Analisis Spasial Temporal Neraca Air DAS Kapuas, Kalimantan Barat Menggunakan Google Earth Engine dan Data Terraclimate

Mohamad Rifai — 2025-12-31

The Kapuas River Basin (DAS Kapuas) is the longest river system in Kalimantan. It plays a crucial role in water resource management, the sustainability of tropical rainforest ecosystems, and regional food security. In the face of climate change pressures, a deep understanding of water balance dynamics is key to effective water resource planning. This study presents a spatial-temporal analysis of the water balance in the Kapuas River Basin over the period 2014–2024, utilizing TerraClimate satellite data and the Google Earth Engine (GEE) cloud computing platform. The main variables such as precipitation, evapotranspiration, and runoff are integrated to calculate the water balance at a monthly scale. Results reveal strong seasonal patterns: average precipitation ranges from 150 to 400 mm/month, peaking between December and February, while evapotranspiration increases toward the end of the dry season (August–October). Runoff shows a high correlation with precipitation, particularly in steep upstream areas. Overall, the Kapuas River Basin is in a state of water surplus; however, significant reductions occurred during El Niño events (2015, 2019, and 2023), indicating vulnerability to climate anomalies. Spatial analysis reveals heterogeneous distribution: upstream and midstream regions exhibit high surplus (>100 mm/month), whereas downstream areas and degraded peatlands show low surplus or mild deficits. The integration of TerraClimate data and GEE proves effective for efficient, large-scale water balance monitoring and is highly replicable, offering a robust foundation for water resource management policies.

Studi Pengaruh Perubahan Tutupan Lahan Akibat Pembangunan Kawasan Perumahan Terhadap Limpasan Permukaan di Kota Pangkalpinang

Rizki Oktian — 2025-12-31

Kota Pangkalpinang mengalami pertumbuhan penduduk yang signifikan. Hal ini mendorong peningkatan pembangunan kawasan perumahan, sehingga mengakibatkan perubahan tutupan lahan dari vegetasi menjadi nonvegetasi yang berdampak pada peningkatan debit limpasan permukaan, risiko banjir, dan penurunan kualitas lingkungan.

Penelitian ini menggunakan metode analisis spasial berbasis citra satelit dari tahun 2004 dan 2023 untuk mengidentifikasi perubahan tutupan lahan akibat pembangunan kawasan perumahan. Pengolahan data dilakukan menggunakan perangkat lunak ArcGIS. Debit limpasan permukaan dihitung dengan metode rasional yang mempertimbangkan koefisien aliran tahun 2004 dan 2023, intensitas hujan, dan luas wilayah. Dampak perubahan tutupan lahan dinilai melalui perubahan nilai koefisien aliran (C) yang selanjutnya digunakan dalam analisis debit banjir rencana menggunakan metode rasional. Hasil analisis menunjukkan bahwa pembangunan kawasan perumahan di Kota Pangkalpinang dalam kurun waktu 20 tahun terakhir mengakibatkan peningkatan rata-rata nilai C, yaitu dari 0,182 menjadi 0,690. Hal ini berkontribusi pada peningkatan debit limpasan permukaan sebesar 285,346%. Penelitian ini menegaskan pentingnya penerapan kebijakan pengelolaan tata ruang berbasis prinsip keberlanjutan, seperti Koefisien Dasar Hijau (KDH) dan pengaturan tata guna lahan, untuk mengurangi dampak negatif perubahan tutupan lahan terhadap kondisi hidrologi perkotaan.

Strategi Penanganan Darurat Groundsill Srandakan dengan Integrasi Model Numerik dan Metode Analitis Shields–Izbash

Sri Wahyuni — 2025-12-31

Sungai Progo, yang berhulu di Gunung Sindoro dan bermuara di Samudera Hindia, memiliki peran penting dalam menopang infrastruktur strategis dan ekosistem di Daerah Istimewa Yogyakarta. Namun, aktivitas vulkanik Gunung Merapi tahun 2010 yang menghasilkan banjir lahar telah meningkatkan suplai sedimen secara signifikan, diikuti dengan aktivitas penambangan galian C yang tidak terkendali. Akumulasi kondisi tersebut mempercepat degradasi morfologi sungai, memicu penurunan dasar sungai hingga 5 meter dalam kurun waktu 20 tahun. Dampak serius dari kondisi ini terlihat pada kerusakan Groundsill Srandakan pada Januari 2025, yang berpotensi mengancam stabilitas jembatan, bendung, serta infrastruktur sungai lainnya. Studi ini bertujuan merancang penanganan darurat Groundsill Srandakan melalui pendekatan model numerik dan metode analitis. Pemodelan numerik dua dimensi menggunakan perangkat lunak HEC-RAS diterapkan untuk menganalisis karakteristik aliran, sedangkan kebutuhan material timbunan ditentukan melalui metode Shields dan Izbash. Hasil analisis menunjukkan kecepatan aliran pada lokasi groundsill mencapai 7 m/s, sehingga dibutuhkan timbunan batu boulder dengan berat minimal 2,2 ton per unit agar stabil terhadap tipe aliran. Rekomendasi desain timbunan ini menekankan efektivitas konstruksi sekaligus menjaga keberlanjutan ekosistem sungai. Penanganan darurat Groundsill Srandakan diharapkan tidak hanya melindungi infrastruktur vital, tetapi juga mendukung konservasi sumber daya air melalui upaya pengendalian degradasi dasar sungai serta pemeliharaan ekosistem perairan.

Integrasi Konservasi SDA dengan Infrastruktur Pengendali Banjir Perkotaan: Studi Kasus Pembangunan Underpass Simpang Joglo, Surakarta

Muhammad Rizky Devianto — 2025-12-31

tantangan baru dalam pengelolaan banjir akibat perubahan tata guna lahan dan sistem drainase. Analisis hidrolik menggunakan HEC-RAS dilakukan untuk mengevaluasi efektivitas jaringan drainase eksisting dan merumuskan alternatif solusi. Tiga skenario simulasi diuji: kondisi eksisting, pemisahan aliran setelah pembuatan talang, dan pemisahan aliran dengan penambahan kolam retensi dan pompa. Hasil studi menunjukkan bahwa skenario ketiga merupakan yang paling efektif, dengan rata-rata penurunan muka air sebesar 0,25 m dan tinggi jagaan (freeboard) sebesar 33 cm. Temuan ini menegaskan bahwa kolam retensi berperan signifikan tidak hanya dalam menyimpan sementara limpasan yang terjadi ketika hujan tetapi juga dalam mendukung infiltrasi untuk mengembalikan air ke tanah dan menjaga kualitas air banjir. Selain perbaikan teknis, pengendalian banjir berkelanjutan di wilayah perkotaan memerlukan integrasi perencanaan tata ruang, konservasi sumber daya air, dan partisipasi masyarakat. Pendekatan ini diharapkan dapat meningkatkan ketahanan infrastruktur sekaligus meminimalkan risiko banjir di masa mendatang.

Penelusuran Waduk Mrica untuk Analisis Perubahan unjuk Kerja Fungsi dan Efisiensi Penggelontoran Sedimen

Shakti Rahadiansyah — 2025-12-31

Mrica Reservoir, also known as the Panglima Besar Soedirman Dam, was constructed in 1982 and serves multiple purposes, including electricity generation, irrigation, and flood control. The main problem faced over the past decade is the very active sedimentation of the reservoir, which is felt to have caused a decline in the reservoir's function as originally planned. This paper presents the results of evaluating the efficiency of sediment removal from Mrica Reservoir by flushing. The RESROUT2 (Reservoir Routing Version 2) was developed to study the water (and sediment) balances and to evaluate the influence of changes in the characteristics of the Mrica Reservoir (with a specific reservoir inflow characteristic) on the outflow pattern. Reservoir outflow is made possible by regulating the output facilities, such as spillways, intake structures for power and irrigation, power generation, and a drawdown culvert to release sediment. The volume of flow out through the bottom outlet and the flushing efficiency for 5 hours are 4.354 million m³ and 0.167, respectively, while for 3 hours of flushing, they are 2.476 million m³ and 0.221, respectively. Increasing the efficiency of sediment flushing in the Mrica Reservoir still needs to be pursued along with efforts to reduce its negative impacts.

Integrating Machine Learning With Geographic Information Systems and Remote Sensing for Erosion Risk Mapping in The Tamalate Watershed

Muhammad Ramdhan Olii — 2025-12-31

Soil erosion poses a significant threat to environmental sustainability, particularly in regions with complex topographic and hydrological characteristics. Accurate erosion risk mapping is essential for effective land management and mitigation strategies. This study aims to evaluate the performance of five machine learning models—Random Forest (RF), Gradient Boosting Tree (GBT), Decision Tree (DT), Generalized Linear Model (GLM), and Support Vector Machine (SVM)—in predicting erosion risk using remote sensing-derived indices. Eight environmental variables, including topographic, hydrological, and vegetation indicators, were analyzed after confirming no harmful multicollinearity (VIF < 3). Model performance was assessed using metrics such as accuracy, AUC, precision, recall, and F-measure. Results show that RF achieved the highest predictive accuracy (0.727) and AUC (0.772), with topographic wetness index (TWI) and normalized difference moisture index (NDMI) being the most influential variables. Conversely, DT tended to overestimate high-risk areas due to overfitting, while SVM and GBT provided more balanced classifications. The spatial classification outcomes revealed that model structure significantly impacts risk distribution, with ensemble models offering more reliable results. Although recall and sensitivity were high across models, specificity was generally low, particularly in GLM and DT, indicating difficulty in detecting non-risk areas. The study highlights the importance of selecting appropriate machine learning approaches and integrating diverse environmental indicators. Future research should address class imbalance and incorporate additional biophysical and socio-economic variables to enhance model robustness and policy relevance.

Analysis of Sedimentation in the Mongiilo River Using the HEC-RAS Program

Syukron Indarwati — 2025-12-31

The river flow that enters the dam has great potential in transporting large amounts of sediment from land erosion that occurs upstream. Sediment material that dissolves with the river flow will then settle in the dam as long as the water structure is in operation. Sedimentation occurs through the process of deposition of soil material which is transported by water or wind. Sedimentation in the riverbed are identified as the causing siltation of river body, irrigation canals, river mouths downstream, and can damage the cross section of the river and have an impact on the effective age of water buildings in the river, in this case the Bulango Ulu Dam. The research location is at the Mongiilo River, Owata Village, North Bulango District, Bone Bolango Regency. This study aims to analyze the sedimentation rate that occurs in the Mongiilo River using the HEC-RAS program. The data used in this study are primary and secondary data. The primary data used for this research was carried out by means of field surveys for sampling, while the secondary data used was obtained by means of literature review and interviews with related agencies such as the Balai Wilayah Sungai Sulawesi II. The results of the analysis in this study obtained that the sediment characteristics in Mongiilo River according to American Geophysical Union are dominated by medium sand, and the sediment transport rate obtained and close to the project data with a value of 200.674,83 tons/year which is using the Meyer-Peter Muller method is 175.484,796 tons/year.

Backflow Analysis in the Downstream of Bolango River through Numerical Integration Method

Namira Salsadilla — 2025-12-31

Rivers are graphic phenomena that form naturally where water flows continuously to lower areas. The downstream or estuary is the end of the river flow which leads to the sea, lake, ocean, or even to a larger river. High-intensity rainfall can lead to overflow because it increases the discharge of water flow in the river beyond the capacity. The overflows that occur downstream and turn upstream are called backflow. This study aimed to examine the occurrence of backflow in downstream of the Bolango River. The research utilized secondary data, such as rainfall and tidal data, and the Bolango-Bone watershed map. Those data were analyzed through hydrological method to obtain design of flood discharges that were collaborated with hydraulics analysis using the Numerical Integration Method to obtain the water level profiles. The results of the hydrological analysis obtained: the design of flood discharge using the Nakayasu Synthetic Unit Hydrograph Method for a 25-year return period was 2046,81 m³/second, a 50-year return period was 2135,72 m³/second, and a 100-year return period was 2209,06 m³/second, and each occurred at the 8th hour. The result of the hydraulics analysis disclosed that the downstream of the Bolango River belonged to the type of sloping surface flow profile or M (Mild) profile because the riverbed slope was smaller than the critical slope.