Bambang Hendro Trisasongko

Despite its significant economic value, tea cultivation in Indonesia is experiencing a continuous decline in plantation area, primarily due to changes in land use. Soil fertility degradation and reduced yields present a significant threat to the long-term sustainability of tea production. A comprehensive understanding of the dynamic nature of tea plantation soil properties is essential for developing sustainable land management strategies aimed at enhancing and maintaining the productivity and sustainability of healthy tea cultivation. This study investigates the properties of Andosols formed from andesite tuff in a heavily managed tea plantation in Cisarua, Bogor, West Java, Indonesia. Four vertical soil profiles and ten composite soil samples were collected from depths of 0-20 cm and 20-40 cm across varying tea vigor and slope class gradients in block plantations. The ongoing production of tea on Andosols has led to significant chemical decline, evidenced by decreased organic matter, increased acidity, reduced cation exchange capacity, and compromised andic properties. The observed rise in base saturation primarily indicates a reduction in CEC, rather than an improvement in fertility. The observed patterns indicate progressive soil weathering and reduced resilience in monoculture systems. Restorative management, which encompasses the incorporation of organic matter and a balanced nutrient supply, is essential for maintaining soil functionality and securing long-term tea productivity. This article synthesizes key findings regarding soil properties, anthropogenic impacts, and strategies for sustainable management. Understanding these dynamics is crucial for optimizing good soil management practices and enhancing tea productivity in volcanic areas.

Mangroves play a critical role in global carbon sequestration, biodiversity conservation, and climate change mitigation. Accurately quantifying mangrove biomass is essential for sustainable forest management and carbon accounting. Yet, the structural complexity and species diversity of mangrove ecosystems pose significant challenges for accurate estimation. In this study, we developed an integrated model that combines multispectral imagery and radar data. Using Sentinel-2 and ALOS-2 satellite imagery combined with field measurements, these data were used to construct linear regression and random forest models for the Matang Mangrove Reserve, Malaysia. We further analyzed the relationships between vegetation indices, radar polarization modes, and biomass. Results indicate that the average biomass is approximately 146 t/ha. The Optimized Soil-Adjusted Vegetation Index (OSAVI) and horizontal–vertical (HV) polarization showed the strongest correlation with field-measured biomass, with an R² of 0.735 and a root mean square error (RMSE) of 46.794 t/ha. This study provides a scientific basis and technical support for mangrove carbon stock assessment, ecosystem management, and climate change mitigation strategies, and highlights the potential of integrating optical and radar remote sensing for large-scale mangrove biomass monitoring.

The potassium (K) Quantity-Intensity (Q-I) relationship results in important parameters, including the activity ratio of potassium at equilibrium (AReK), which indicates potassium availability in soil. Experiments to observe soil Q-I K relationship parameters are often complex, time-consuming, and do not include environmental variables. This research aims to model AReK using a machine learning (ML) approach. ML models applied are Random Forest (RF), Cubist, and Support Vector Machine (SVM) as the primary approaches, with Multiple Linear Regression (MLR) serving as a baseline. The dataset was derived from sixty-one observation points in Brebes, Central Java. The predictors were pH, organic carbon, clay, cation exchange capacity (CEC), exchangeable cations (Exc-Ca, Mg, K, Na), water soluble K, available K, K saturation, potential K, non-exchangeable K (NE-K), elevation, and slope. The response variable was the AReK. Variable selection was performed using Pearson correlation to eliminate highly correlated predictors and reduce multicollinearity. Exactly 75% of the data was utilized as the training set and 25% as the test set. Three metrics, i.e., MAE, RMSE, and R², were used in model evaluation. The results showed that the Cubist model could predict AReK with high accuracy (R2=0.9437) and low RMSE (0.5701) and MAE (0.3514). Based on the Cubist model, Exc-K, Exc-Mg, CEC, and Exc-Ca were the most important variables for predicting AReK. This model can be employed to support site-specific fertilizer recommendation strategies. To improve the performance of the model, it is necessary to add other predictor variables (e.g., soil physical properties, clay minerals, rainfall, temperature and soil moisture).

Purpose. This study examines how climatic factors such as land surface temperature, precipitation, relative humidity, solar radiation, soil moisture, and the Index of Prices Paid by Farmers (Ib) influence rice production in Banten Province, Indonesia. The study used panel data from 2008 to 2024 to analyse these effects. Results. Among all models tested, the Random Forest algorithm demonstrated the highest predictive performance in modelling rice production in Banten Province. It achieved the lowest Root Mean Square Error (RMSE) of 0.239 and the highest coefficient of determination (R² = 0.758), indicating that the model could explain approximately 75.8% of the variation in rice yields. In contrast, traditional linear regression produced lower accuracy (R² = 0.559; RMSE = 0.312). The Random Forest model effectively captured nonlinear interactions and multicollinearity between climatic and economic variables. Variable importance analysis revealed that land surface temperature and relative humidity were the most influential predictors, while precipitation and solar radiation had moderate effects. The Index of Prices Paid by Farmers (Ib), although statistically insignificant in the linear model, showed a nonlinear threshold effect in the Random Forest model, negatively affecting production beyond certain levels. Accumulated Local Effects (ALE) plots further confirmed these nonlinear and threshold-based relationships, visually interpreting how each predictor influenced rice output across its distribution. These findings underscore the capability of machine learning not only to enhance predictive accuracy but also to interpret variable importance in a nuanced manner. Scientific novelty. This study offers a methodological advancement by integrating climatic and economic variables into a comprehensive predictive framework using ensemble machine learning models. While previous studies often relied solely on linear models or focused only on climatic data, this research includes the Index of Prices Paid by Farmers (Ib) to reflect input cost pressures, a factor rarely explored in similar contexts. Moreover, Accumulated Local Effects plots enhance model interpretability by illustrating the marginal effects of predictors, overcoming common limitations of black-box machine learning models. The study’s novelty lies in its dual contribution: achieving high predictive accuracy and delivering interpretable insights relevant to agricultural planning in climate-sensitive regions. Practical value. The findings provide critical evidence to support the formulation of climate-resilient and economically sensitive agricultural policies. The strong influence of temperature and humidity implies an urgent need to invest in adaptive technologies such as heat-tolerant rice varieties and region-specific irrigation systems. Furthermore, identifying a cost threshold in the Ib variable suggests that rice production may decline beyond certain input price levels, highlighting the importance of timely and targeted subsidies, input price control, or financial support mechanisms for smallholder farmers. The machine learning framework developed in this study can also be adopted as a decision-support tool for early warning systems, helping policymakers predict potential yield shortfalls and plan proactive interventions to maintain food security under uncertain climatic conditions.

Multispectral sensors have been invaluable for land cover mapping. Nonetheless, cloud cover is frequent in many parts in the tropics, meaning that relying on single-date multispectral data can be problematic. The limited number of variables, i.e., only two, in conventional dual-polarized synthetic aperture radar data usually drives low classification accuracies. Researchers have explored an augmentation strategy that uses textural features, such as Haralick’s Gray Level Co-occurrence Matrix and/or decomposition features. However, this also does not produce an acceptable outcome. Another approach involves short-term temporal stacking. This masks out atmospheric disturbance and means every pixel has a single value representing the object. Our research explored this last approach to identify woody vegetation in a rural area of Indonesia from Sentinel-2 data. Ten machine learning algorithms were explored to investigate limitations and opportunities. We investigated algorithms capable of reporting variable importance to indicate crucial spectral bands. This study indicated that highly important spectra include red edge and short-wave infrared bands and to a lesser extent, the red band. Part of these bands were confirmed by the variable importance. The spectral similarity between bands 8 and 8A could be seen as a drawback when both bands are employed. Variable importance indicated a negative impact due to this similarity, suggesting that only one should be used in the classification. Overall, classification accuracy was generally high (over 96%), except in the case of the benchmark, i.e., classification and regression trees. Highest accuracy (97.6%) was provided by C5.0 model, although the difference from random forests (RFs) was negligible. The RF was favored as it produced a better estimate of timber plantations. A few algorithms were unable to separate woody vegetation types, including the popular support vector machine. Successful separation was mainly achieved by ensemble tree learners, such as RFs and the variants of extreme gradient boosting.