Balancing water supply demand is vital for sustaining livelihoods. Spatial mapping and calculating water yield dynamics due to land use changes over decades are needed to manage land resources and formulate ecological protection policies. This study mapped the supply, demand, and matching status of water product service using the Integrated Valuation of Ecosystem Service and Tradeoff (InVEST) biophysical models in the Citarum Watershed (CW) in 2000, 2010, and 2020. Moreover, this study used Exploratory Spatial Data Analysis (ESDA) and Geographic Information System (GIS) techniques to study the agglomeration characteristics and evolutionary trajectories of supply–demand over two decades. The results showed that between 2000–2010 and 2010–2020, the water supply decreased by 19.01 × 10⁸ m³ (18.28%) and 12.97 × 10⁸ m³ (15.27%), respectively. However, the water demand in the same period increased by 6.17 × 10⁸ m³ (23%) and 15.74 × 10⁸ m³ (47%), respectively. Over the decades, the contribution of land use land cover (LULC) changes to variations in water supply has yielded values ranging from 2.87% to 6.37%. The analysis of the water supply–demand imbalance indicated that the entire CW experienced water shortage, and the type of spatial matching for supply and demand is dominated by a high supply and high demand class (16.09% of the total area). Based on the level of water deficit calculation, the upstream and downstream areas were identified as zones that require ecological conservation, while the middle CW area requires ecological restoration or ecological improvement.

The application of remote sensing data has been significant in modeling soil erosion. However, previous studies have fallen short in elucidating and lacked an understanding of the multifactor influencing erosion. This study addresses these limitations by employing the InVEST and the Geodetector models. Specifically, it aims (1) to delineate both spatial and temporal variations in soil erosion within the Citarum watershed from 2010 to 2020, (2) to identify the key drivers of soil erosion and unravel the underlying mechanisms, and (3) to identify the high-risk zones for soil erosion. Both models consider a range of natural predictors, including topography (slope factor), climate (precipitation factor), and vegetation cover (vegetation factor). In addition, they incorporate social parameters such as income per capita and population density, which interact with the watersheds position in the downstream, middle, and upper streams. The results reveal that, over a decade, the average soil erosion increased by 15.50×10⁶ tons, marking a 16.65 surge. The impact of factors varies significantly across different subwatershed areas. For example, fraction vegetation cover interactions influence upper- and middle-stream regions, while the downstream area is notably affected by precipitation interactions. The high-risk erosion areas in the watershed are primarily influenced by slope, precipitation, and fractional vegetation cover. In these areas, factors causing high erosion risks include slope, precipitation, and other environmental variables categorized into strata. The study highlights the varying influential factors in different watershed areas.

Soil is a solid-particle that covers the earth's surface. Soils can be classified based their color. The color can be an indication of soil properties and soil conditions. Soil image classification requires high accuracy and caution. CNN works well on image classification, but CNN requires a large amount of data. Augmentation is one technique to overcome data needs like rotation and improving contrast. Rotation is the movement of rotating the image position randomly to various degrees. Gamma Correction is a method to improve image by decreasing or increasing the contrast. The rotation and Gamma Correction on augmentation can increase the amount of training data from 156 to 2500 soil images data. The classification of soil data is not referred to soil taxonomy system such as Entisols and Histosols but it used arbitrary simple classification based on color. Unfortunately, the weakness of the CNN is vanishing and exploded gradients. Another Deep learning that can overcome vanishing and exploded gradients is dense blocks. This study proposes a combination of Augmentation and CNN-Dense block where in the augmentation a combination of rotation and Gamma-correction techniques is used and Soil image classification based on color is used by the CNN-Dense block. The combination method is able to give excellent results, where all performances accuracy, precisions, recall and F1-Score are above 90%. The combination of rotation and Gamma Correction on augmentation and CNN is a robust method to use in soil image classification based on color.