Deep learning (DL) approaches have demonstrated the potential for accurate groundwater level (GWL) forecasting. However, the need to train the DL model independently for each station poses a significant challenge for large-scale applications. This study explored the potential of a transfer learning (TL) approach coupled with long short-term memory (LSTM) (TL-LSTM) networks for basin-wide GWL forecasting. The performance of TL-LSTM was compared against independently trained LSTM (IT-LSTM) models, which showed superior accuracy over several machine learning models for Kumamoto, Japan. The basin average values of performance indicators such as the coefficient of determination (R2) and Nash-Sutcliffe efficiency (NSE) for IT-LSTM models were 0.93, 0.86, and 0.78 (R2) and 0.91, 0.86, and 0.77 (NSE) for 1-, 2-, and 3-month-ahead forecasting, respectively. Similar performances were achieved using the new model with significantly less training effort and computational time, making it suitable for large-scale GWL forecasting.

Surface irrigation is vital for maintaining agricultural productivity and preventing regional food insecurity. The current study aims to identify potential surface irrigation sites in the Koch Bihar district via an innovative geospatial-based E-PROMETHEE II multi-criteria decision-making (MCDM) technique. Fourteen factors, including distance from water bodies, river density, rainfall, temperature, evapotranspiration, the normalized difference vegetation index (NDWI), the soil moisture index, elevation, slope, aspect, soil properties, groundwater depth and land use/land cover, were selected for analysis, with a multicollinearity test ensuring their validity. The E-PROMETHEE II method was used to create a suitability map for surface irrigation, benchmarked against the analytical hierarchical process (AHP) method. The results highlighted highly suitable zones for irrigation, notably in the Cooch Behar I, Dinhata I, Tufanganj I, Mathabhanga II, and parts of the Tufanganj II blocks. Validation via receiver operating characteristic (ROC) and area under the curve (AUC) analyses demonstrated AUC values exceeding 0.70 for all the models, underscoring their reliability. E-PROMETHEE II (AUC = 0.889) emerged as the top-performing method, trailed by the AHP (AUC = 0.728). These findings will facilitate the implementation of new surface irrigation systems in the study area. L'irrigation de surface est vitale pour maintenir la productivit & eacute; agricole et pr & eacute;venir l'ins & eacute;curit & eacute; alimentaire r & eacute;gionale. La pr & eacute;sente & eacute;tude vise & agrave; identifier les sites potentiels d'irrigation de surface dans le district de Koch Bihar via une technique innovante de prise de d & eacute;cision multi-crit & egrave;res (MCDM) bas & eacute;e sur les donn & eacute;es g & eacute;ospatiales E-PROMETHEE II. Quatorze facteurs, dont la distance par rapport aux corps d'eau, la densit & eacute; des cours d'eau, les pr & eacute;cipitations, la temp & eacute;rature, l'& eacute;vapotranspiration, l'indice de v & eacute;g & eacute;tation par diff & eacute;rence normalis & eacute;e (NDWI), l'indice d'humidit & eacute; du sol, l'& eacute;l & eacute;vation, la pente, l'aspect, les propri & eacute;t & eacute;s du sol, la profondeur des eaux souterraines et l'occupation/la couverture des sols, ont & eacute;t & eacute; s & eacute;lectionn & eacute;s aux fins d'analyse, un test de multicollin & eacute;arit & eacute; assurant leur validit & eacute;. La m & eacute;thode E-PROMETHEE II a & eacute;t & eacute; utilis & eacute;e pour cr & eacute;er une carte d'aptitude & agrave; l'irrigation de surface, par rapport & agrave; la m & eacute;thode AHP. Les r & eacute;sultats ont mis en & eacute;vidence des zones tr & egrave;s appropri & eacute;es pour l'irrigation, notamment dans les blocs de Cooch Behar I, Dinhata I, Tufanganj I, Mathabhanga II et certaines parties des blocs de Tufanganj II. La validation par les analyses des caract & eacute;ristiques de fonctionnement du r & eacute;cepteur (ROC) et de l'aire sous la courbe (AUC) a d & eacute;montr & eacute; des valeurs AUC sup & eacute;rieures & agrave; 0,70 pour tous les mod & egrave;les, ce qui souligne leur fiabilit & eacute;. E-PROMETHEE II (AUC = 0,889) est apparu comme la m & eacute;thode la plus performante, suivie par le processus hi & eacute;rarchique analytique (AHP) (AUC = 0,728). Ces r & eacute;sultats faciliteront la mise en oe uvre de nouveaux syst & egrave;mes d'irrigation de surface dans la zone d'& eacute;tude.

Contemporarily, among the crucial techniques for generating precise land suitability map site suitability analysis (SSA) is the most acceptable one. This chapter intended to determine the fittest areas for producing potato crops in the Jalpaiguri district of West Bengal, India. Twenty-four criteria were selected for the study, and their weights were determined using the GIS-based fuzzy analytical hierarchy processFuzzy analytical hierarchy process (F-AHP) approach and expert judgment. Based on the results, soil texture (0.1349), soil organic carbon (0.0614), geomorphology (0.1523), slope (0.1043), soil moisture index (0.0845), temperatureTemperature (0.0769), nitrogen (0.0529), and soil pH (0.0432) and rainfallRainfall (0.0635) are among the most crucial criteria for producing SSA for potato production. It has been observed that regions for potato production are extremely favorable (15.30%), highly suitable (25.11%), moderately suitable (28.55%), marginally suitable (22.52%), and not suitable (8.52%), and their geographical extend covers 508.83 km2, 835.05 km2, 949.35 km2, 749.04 km2, and 283.43 km2 area, respectively. The area confirms the F-AHP approach's reliability in SSA under curve (AUC), and the value of AUC is 0.83. The results also have been verified with the help of GPS and an interview survey from the potato cultivators. The studies indicate expected scenarios for agricultural activity. Thus, an appropriate sustainable plan should be prepared to increase the regional agrarian output, and the local farmersFarmers should use targeted adaptable farming practice(s). Farmers, regional planners, and government representatives may utilize the proposed map, containing information about the agricultureAgriculture suitability, to make essential choices for the area, including identifying viable agricultural areas, promoting agricultural development, and supporting local independent companies for potato farming.

Land Suitability Analysis (LSA) is one of the most important methods for developing a precise land suitability map. The present study was aimed to identify suitable paddy crop-growing regions in the Koch Bihar district of West Bengal, India. For the study, 18 criteria were chosen, and their weightage was assigned based on expert opinion and the GIS-based Fuzzy Analytical Hierarchy process (F-AHP) method adopted. Results indicated the most important factors for developing LSA for paddy cultivation are rainfall (16.72%), soil texture (13.22%), slope (11.55%), elevation (09.63%), soil moisture (08.27%), temperature (07.11%), and pH (05.03%). It has been observed that highly suitable (14.54%), moderately suitable (46.07%), marginally suitable (24.20%), and not suitable (15.19%) areas for paddy cultivation. The Area Under Curve (AUC) value (0.869) confirms the reliability of the F-AHP approach in LSA. The findings of the study provide sustainable agricultural planning and delineate potential sites for paddy cultivation.

The present study was designed to assess the capability of the novel Multiple-criteria decision-making (MCDM) method such as Technique for Order of Preference by Similarity to Ideal Solution (TOPSIS) along with Analytical Hierarchy Process (AHP) in identifying the potential site of ecotourism in the Eastern Dooars region of West Bengal, India. Two popular objective weighting methods such as Entropy and Criteria Importance through Interaction Correlation (CRITIC), were used to minimize the subjectivity of AHP methods. The suitability map obtained from these applied methods showed that the northern and north-eastern parts, namely Kalchini and Kumargram blocks, were identified as the highly suitable zone for ecotourism development. Finally, the obtained suitability map was validated using receiver operating characteristic (ROC) and Area Under Curve (AUC) analysis. The AUC value of ROC was more than 0.80 for all the applied methods, indicating the reliability of all the applied models.

Climate change has posed a serious threat to agriculture over the last few decades. It has become a source of concern for a country like Bangladesh, which is heavily reliant on agriculture. Therefore, it is necessary to understand possible changes in climatic variables and their impact on agriculture on a microscale. The present study was designed to assess the farmer's perception of climate change in Rajshahi and Bogra Districts of Bangladesh. This study also investigated whether farmers' perceptions of climate change were consistent with climatic records. Furthermore, the study investigated the selection of various adaptation strategies taken by farmers to cope with climate change and mitigate its effects. The annual rainfall trend from the MK/mMK test portrayed that most stations experienced decreased rainfall. Similar results were also achieved on a seasonal scale. In contrast, the annual and seasonal temperature trends had shown an increase in most of the stations. The results were perfectively in line with the farmers' perception. Results also showed that farmers had adopted four broad strategies to deal with the climate change scenario viz. technological development, government programs, farm production, and financial management. However, the dominant strategies adopted by farmers were changes in planting dates, crop rotation, supplementary irrigation, and fertiliser changes. Overall, the study's findings highlight the need for immediate actions to mitigate the negative impacts of climate change and help develop sustainable adaptation strategies to maintain agricultural productivity.

Accurate drought forecasting is necessary for early warning of drought hazards, water resources and eco-environmental management. This study assessed the capability of the Prophet model to forecast meteorological drought based on the standardized precipitation index (SPI). A comparative assessment was conducted among Prophet, support vector regression (SVR) and multiple linear regression (MLR) models to test the applicability of the new model. The SPI was computed at multiple time scales (SPI3, SPI6, SPI12 and SPI24) for 38 meteorological stations located in semi-arid areas in western India. Minimum redundancy maximum relevance (mRMR) was applied to select input variables prior to the model development. The results reveal that the Prophet model yielded acceptable accuracy for drought forecasting, whereas SVR and MLR models showed greater error for short-term forecasting in terms of coefficient of determination (R2) and Nash-Sutcliffe efficiency (NSE). Therefore, the Prophet model is recommended as a new robust model for drought forecasting.

Visualization of present state of aquifers and identification of groundwater depletion hotspots are important tools in preparing an effective groundwater management plan. Therefore, this study developed an integrated framework by bridging a number of relevant factors to characterize and visualize groundwater depletion hotspots in Andhra Pradesh, India. Firstly, the groundwater status was assessed by detecting spatio-temporal trends in groundwater levels of 429 dug well sites from 2004 to 2018 using Mann-Kendall (MK)/modified Mann-Kendal (mMK), Spearman’s Rho test, and the magnitude of the slope was determined by Sen’s slope estimator. Subsequently, multiple decision factors were considered in the analytical hierarchy process (AHP) method for producing the groundwater stress zone map. A multicollinearity test was performed prior to the incorporation of these factors in order to improve the decision-making power of the AHP method. The results of the groundwater stress zoning map showed that 19.99%, 16.93%, 24.63%, 18.86% and 19.59 % of areas were classified as low, moderate, high and very high stress zones, respectively. Results also identified the south-western parts as groundwater depletion hotspots. Furthermore, validation results using Sen’s slope map, evaluation metrics of ROC (receiver operating characteristics) and AUC (area under curve) showed that AHP method had exhibited a reliable performance with an accuracy of 76.7%. Thus, the applied integrated approach can be used to explicitly characterize groundwater status by integrating different factors. The findings of our study also would be helpful for water resources managers and planners who need to design proper and sustainable management of groundwater resources.

Groundwater level (GWL) forecasting is crucial for irrigation scheduling, water supply and land development. Machine learning (ML) (e.g., artificial neural networks) has been increasingly adopted to forecast GWL due to its ability to model nonlinearities between GWL and its drivers (e.g., rainfall). Although ML approaches have been successful at forecasting GWL, they are often inaccurate when GWL exhibits multiscale changes (e.g., due to urbanization). To address this shortcoming, wavelet transforms (WT) are routinely coupled with ML methods. Unfortunately, researchers frequently neglect key issues associated with WT that render such forecasts useless for real-world scenarios. This study demonstrates how new ML methods, such as eXtreme Gradient Boosting and Random Forests, can be properly coupled with WT to generate accurate GWL forecasts (1-3 months ahead) for 7 wells in Kumamoto City in Southern Japan that can be used to help address current pressing issues such as groundwater quality and land subsidence.