Agricultural Planning Research Articles

Explainability in wind farm planning: A machine learning framework for automatic site selection of wind farms

The process of wind farm site selection involves complex considerations spanning environmental, economic, and social factors, often handled through multi-criteria decision making (MCDM) methods. However, MCDM approaches face several challenges. To address these, integrating machine learning (ML) techniques offers promise. Nevertheless, existing studies have yet to employ a standalone ML-based approach for this purpose. This study seeks to pioneer an explainable ML-based framework for automated wind farm site selection. The framework incorporates feature selection techniques, seven ML algorithms, statistical tests, and a local explainability method. The experimental area was chosen as Balıkesir province, which hosts 13% of Türkiye's installed wind energy capacity. According to the prediction performances of the ML models, Extreme Gradient Boosting-XGBoost has the highest accuracy (0.9607), followed by Light Gradient Boosting Machine-LightGBM (0.9580), Random Forest-RF (0.9518), Histogram-based Gradient Boosting-HGB (0.9387), Classification and Regression Trees-CART (0.8946), Logistic Regression-LR (0.8856), and Naive Bayes-NB (0.8456). Additionally, McNemar’s tests revealed statistically significant differences among ML models. Based on the explanations provided by SHapley Additive exPlanations, wind speed, distance to transmission lines, distance to protected zones, and elevation were the top contributing criteria. Moreover, analysis revealed regions with elevated wind speeds, higher elevations, and closer proximity to transmission lines and protected areas as favorable for wind farm installation. The study's outcomes indicate strong agreement (97%) between current wind turbine locations and those identified as suitable by ML models. Within Balıkesir province, our research has identified a significant number of highly suitable areas where wind turbines have not yet been installed. These findings underscore the critical importance of these locations for future investments in wind farm development. The proposed framework could be highly efficient for wind farm siting studies in provinces that share similar characteristics with the study area used in this study offering a robust tool for future wind farm siting studies. Finally, the study contributes to achieving UN SDG 7 by employing ML to refine wind farm site selection, thereby facilitating the transition towards sustainable and universally accessible energy.

Flood susceptibility mapping contributes to disaster risk reduction: A case study in Sindh, Pakistan

Floods are a widespread and damaging natural phenomenon that causes harm to human lives, resources, property and has agricultural, eco-environmental, and economic impacts. Therefore, it is crucial to perform flood susceptibility mapping (FSM) to identify susceptible zones to mitigate and reduce damage. This study assessed the damage caused by the 2022 flash flood in Sindh and identified flood-susceptible zones based on the frequency ratio (FR) and analytical hierarchy process (AHP) models. Flood inventory maps were generated, containing 150 flood sampling points, which were manually selected from Landsat imagery. The flood sampling points were split into 70 % for training and 30 % for validating the results. Furthermore, fourteen flood conditioning factors were considered in the analysis for developing FSM. The final FSM is categorized into five zones, representing levels from very low to very high. The results identified that the areas under very high flood-susceptible zones in Ghotki (FR 4.42 % and AHP 5.66 %), Dadu (FR 21.40 % and AHP 21.29 %), and Sanghar (FR 6.81 % and AHP 6.78 %). Ultimately, FSM accuracy was evaluated using the receiver operating characteristics and area under the curve method, resulting in Ghotki (FR 82 %, AHP 83 %), Dadu (FR 91 %, AHP 90 %), and Sanghar (FR 96 %, AHP 95 %). This study enhances the scientific understanding of flash flood impacts across diverse regions and emphasizes the importance of accurate FSM for informed decision-making. The findings provide important informations and are supportive for policymakers, agricultural planners, and stakeholders engaged in flood risk management to mitigate the adverse consequences of flash floods.

A novel method for studying the wind speed probability distribution and estimating the average wind energy density

Traditional distribution models generally have large fitting errors at low wind speeds and poor fitting effects at multi-peak wind speed distributions. In this paper, a novel approach is proposed to fit different wind speed distributions, introducing a Gumbel distribution into common hybrid distribution models. The model parameters are solved by a combination of snake optimizer and nonlinear least squares (SO-NLS), using the optimal values obtained by the nonlinear least squares method as a set of initial input vectors for the snake optimizer. Simulation experiments were conducted using multi-peak wind speed distribution datasets with varying characteristics, comparing the fitting performance of the improved hybrid models against the conventional Weibull, Normal, and Rayleigh hybrid models. The results show that the proposed approach improved the model fit effects, particularly at low wind speeds, in all five experimental datasets. In most cases, the overall fitting effects were also improved. Furthermore, the validity and superiority of the improved hybrid models were further verified by comparing the estimated average wind energy density. Meanwhile, the experimental results also verified that SO-NLS not only yielded better optimization results but also accelerated the convergence speed than the snake optimizer. The improvements presented in this study effectively address the problem of large fitting errors at the low wind speed sections of the distribution, providing a theoretical basis for wind farm planning and design.

WSN-Driven Advances in Soil Moisture Estimation: A Machine Learning Approach

Soil moisture estimation is crucial for agricultural productivity and environmental management. This study explores the integration of Wireless Sensor Networks (WSNs) with machine learning (ML) and deep learning (DL) techniques to optimize soil moisture estimation. By combining data from WSN nodes with satellite and climate data, this research aims to enhance the accuracy and resolution of soil moisture estimation, enabling more effective agricultural planning, irrigation management, and environmental monitoring. Five ML models, including linear regression, support vector machines, decision trees, random forests, and long short-term memory networks (LSTM), are evaluated and compared using real-world data from multiple geographical regions, which includes a dataset from NASA’s SMAP project, supplemented by climate data, which employs both active and passive sensors for data collection. The outcomes demonstrate that the LSTM model consistently outperforms other ML algorithms across various evaluation metrics, highlighting the effectiveness of WSN-driven approaches to soil moisture estimation. The study contributes to the advancement of soil moisture monitoring technologies, offering insights into the potential of WSNs combined with ML and DL for sustainable agriculture and environmental management practices.

Farmer Application

The aim of the program is to provide a platform for farmers to rent modern agricultural machinery and equipment as well as rent storage space for crops. The app will also connect to government servers to track data on crops, fertilizer use and government funds. Records of machinery and equipment will be kept along with details of the crop stored. In India, agriculture is the main source of income for the majority of the population. However, many farmers are not aware of advances in agricultural technology and cannot understand the market value of their products. That's why they often sell products at very low prices. In addition, they also try to follow the government's agricultural plans and reports. To solve these problems, our initiative will create a platform where farmers can get the latest information on vegetables and fruits in every market in India. This will help them sell their products at reasonable prices and improve their financial situation. Additionally, farmers will also receive notifications regarding government agricultural programs and in-app notifications to benefit from these programs. Additionally, our application will help farmers prepare the required products by estimating the amount of vegetables and fruits according to the market. We will also include weather data to support farmers' planning for the next 2-3 days. Overall, our programs focus on providing farmers with new knowledge and technology that will improve their livelihoods and increase their productivity

In-Season Cotton Yield Prediction with Scale-Aware Convolutional Neural Network Models and Unmanned Aerial Vehicle RGB Imagery.

In the pursuit of sustainable agriculture, efficient water management remains crucial, with growers relying on advanced techniques for informed decision-making. Cotton yield prediction, a critical aspect of agricultural planning, benefits from cutting-edge technologies. However, traditional methods often struggle to capture the nuanced complexities of crop health and growth. This study introduces a novel approach to cotton yield prediction, leveraging the synergy between Unmanned Aerial Vehicles (UAVs) and scale-aware convolutional neural networks (CNNs). The proposed model seeks to harness the spatiotemporal dynamics inherent in high-resolution UAV imagery to improve the accuracy of the cotton yield prediction. The CNN component adeptly extracts spatial features from UAV-derived imagery, capturing intricate details related to crop health and growth, modeling temporal dependencies, and facilitating the recognition of trends and patterns over time. Research experiments were carried out in a cotton field at the USDA-ARS Cropping Systems Research Laboratory (CSRL) in Lubbock, Texas, with three replications evaluating four irrigation treatments (rainfed, full irrigation, percent deficit of full irrigation, and time delay of full irrigation) on cotton yield. The prediction revealed that the proposed CNN regression models outperformed conventional CNN models, such as AlexNet, CNN-3D, CNN-LSTM, ResNet. The proposed CNN model showed state-of-art performance at different image scales, with the R2 exceeding 0.9. At the cotton row level, the mean absolute error (MAE) and mean absolute percentage error (MAPE) were 3.08 pounds per row and 7.76%, respectively. At the cotton grid level, the MAE and MAPE were 0.05 pounds and 10%, respectively. This shows the proposed model's adaptability to the dynamic interplay between spatial and temporal factors that affect cotton yield. The authors conclude that integrating UAV-derived imagery and CNN regression models is a potent strategy for advancing precision agriculture, providing growers with a powerful tool to optimize cultivation practices and enhance overall cotton productivity.

Rainfall prediction for data scares areas using metrological satellites in the case of the Lake Tana sub-basin, Ethiopia

Abstract In African nations with complex topographies, alternative rainfall estimation methods such as satellites are crucial. This study is aimed at predicting the spatial and temporal distribution of rainfall in the Lake Tana sub-basin from 1990 to 2020. A satellite-based rainfall estimate of Climate Hazards Group Infrared Precipitation with Station data (CHIRPS) was used with the same spanning period (1990–2020). The validation process employs point-to-pixel analysis, comparing CHIRPS estimates with observed data at specific gauge stations. The findings showed that CHIRPS had well-estimated rainfall incidence in the highland areas and significantly overestimated it in the lowland areas. The Mann–Kendall trends for January, June, and August indicate decreasing trends, while the Bega and spring seasons show notable declines. Regression analysis reveals a non-significant decrease in annual rainfall with the highest rainfall in the summer and relatively dry winters. In addition, the coefficient of variation value of 26.37% suggests a moderate level of variability around the mean annual rainfall. In conclusion, the CHIRPS satellite exhibited varied performance across the Tana Sub-basin, with site-specific discrepancies and notable inaccuracies at certain stations. The study underscores the importance of considering local factors and topography in satellite-based rainfall assessments, providing valuable insights for agricultural planning in the region.

Assessment of Frost Depth and Frost Heave in Romania

The investigation of extreme weather and climate events has gained momentum in the last few decades, spurred in part by political involvement. However, not enough attention has been paid to the historical recorded data of climate monitoring. An accurate estimate of maximum soil frost depth is an important factor in determining construction costs and structures’ foundations, and designers also need reliable information about local meteorological parameters. In this study, the calculations of the parameters for the city of Arad are presented as an example. An updated Romanian frost depth map and a new Romanian freeze–thaw cycle (FTC) map are needed since recent climate studies have documented an increase in global and regional surface temperatures, with the greatest shift in warming occurring over the last three decades. This trend is particularly pronounced when comparing the 1990–2020 period with earlier periods, which could be indicative of broader climate change trends. It is important to note that these conclusions are based solely on the provided statistical parameters and do not take into account other potential factors influencing temperature trends in the region. The results from this investigation could be used to achieve a reduction in frost disasters and the formulation of policies and measures for the adaptation of geotechnical/structural design for Romanian territory. They may also support the development of maps that help to visualize and understand the evolving climate patterns in the region, including changes in frost depth and the frequency of freeze–thaw cycles, which are important for various sectors such as agriculture, construction, and infrastructural planning. Given the documented increase in global and regional surface temperatures, updating such maps will also provide valuable information for policymakers, researchers, and stakeholders on how to adapt to ongoing climate change and its impacts on the Romanian region.

ChinaRiceCalendar – seasonal crop calendars for early-, middle-, and late-season rice in China

Abstract. Long time series and large-scale rice calendar datasets provide valuable information for agricultural planning and field management in rice-based cropping systems. However, current regional-level rice calendar datasets do not accurately distinguish between rice seasons in China, causing uncertainty in crop model simulation and climate change impact analysis. Based on satellite remote sensing data, we extracted transplanting, heading, and maturity dates of early-, middle-, and late-season rice across China from 2003 to 2022 and established a multi-season rice calendar dataset named ChinaRiceCalendar (https://doi.org/10.7910/DVN/EUP8EY, Liu et al., 2023). Overall, the ChinaRiceCalendar dataset shows good agreement with field-observed phenological dates of early-, middle-, and late-season rice in Chinese agricultural meteorological stations (AMSs). According to the calendar data from 2003 to 2022 in China, the transplanting dates for early-, middle-, and late-season rice shifted by +0.7, −0.7, and −5.1 DOY (day of year) per decade, respectively; the heading dates for early-, middle-, and late-season rice shifted by −0.5, +2.7, and −0.6 DOY per decade, respectively; the maturity dates for early-, middle-, and late-season rice shifted by −0.7, +3.8, and −1.6 DOY per decade, respectively. ChinaRiceCalendar can be utilized to investigate and optimize the spatiotemporal structure of rice cultivation in China under climate and land use change.

New approach for forecasting rice and corn production in the Philippines through machine learning models

This study presents a comprehensive analysis of machine learning models for forecasting rice and corn production in the Philippines, focusing on determining the most effective model for this purpose. Given the crucial role of these crops in the nation's economy and food security, accurate forecasting is essential. We compared four different models: Random Forest (RF), Echo State Network (ESN), Neural Network Autoregressive (NNAR), and Autoregressive Support Vector Machine (ARSVM), using historical production data from 1987 to the first quarter of 2023. The Random Forest model, configured with 500 trees and one variable at each split, emerged as the most accurate, demonstrating the lowest error rates across Mean Absolute Error (MAE), Root Mean Square Error (RMSE), and Mean Absolute Percentage Error (MAPE) metrics for both crops. This superior performance can be attributed to RF's ability to capture complex, non-linear relationships in the data. In contrast, ESN, NNAR, and ARSVM showed varying levels of accuracy, with ARSVM recording the highest error rates. This variation underlines the importance of model selection based on the specific characteristics of agricultural data. The findings of the study have significant implications for agricultural forecasting and planning in the Philippines. They highlight the potential of using advanced machine learning techniques to improve crop production predictions, thereby aiding in better resource allocation and policy formulation for sustainable agricultural development.

Smallholder farmers’ perceived motivations for the adoption and implementation of climate information services in the Atwima Nwabiagya District, Ghana

This study investigated smallholder farmers’ perceived motivations for the adoption and implementation of climate information services (CIS) in the Atwima Nwabiagya District, Ghana. Specifically, the study answered the following research questions. (i) What is the perception of smallholder farmers on climate change and variability in the Atwima Nwabiagya District?; (ii) What are the CIS accessed and used by the smallholder farmers for farm management decisions?; (iii) What are the perceived motivations for the adoption and implementation of CIS by smallholder farmers?; and; (iv) What are the barriers confronting smallholder farmers in adopting and implementing CIS? Primary data was collected through questionnaire surveys with 200 farming households and four key informant interviews in four selected communities (Kobeng, Seidi, Wioso, and Nkawie Panin) in the study district. Results showed that smallholder farmers perceived climate change and variability through increased windstorms and thunderstorms, variable rainfall, and increased temperature patterns. The surveyed farmers also perceived that the climatic changes affect their agricultural activities adversely. However, the smallholder farmers accessed and used CIS including weather warnings (86%), daily weather forecasts (70%), and seasonal forecasts (60%) to manage climatic risks on the farm. Results also showed that the smallholder farmers were motivated by critical factors including improvement in the acquisition of proactive measures (92%), increased knowledge of climatic patterns that affect food stock (91%), and a reduction in loss of stock produced (88%) to access and use CIS. Despite this, the smallholder farmers were confronted with key barriers in their uptake and utilization of CIS. These were the timeliness of CIS (73%), a lack of weather symbols in the forecasts (65%), and inadequate information on seasonal forecasts for long-term planning (61%). The study concluded that the motivations for smallholder farmers using CIS in the study district may vary depending on their context, goals, and challenges. However, the general motivation is to enhance smallholder farmers’ farm planning and risk management by using weather and climate forecasts to make better decisions on crop selection, planting dates, input use, pest and disease control, irrigation, harvesting, and marketing.

State information systems as a part of land use planning in agriculture

In the article, the authors discuss the lack of a unified source of information that can be used for solving land use planning problems in agriculture. Highlighting different levels of storage of such information, some state information systems are considered in this paper as the most important for solving such problems. The authors also propose the possible improvements in this sphere.

Socio-Economic and Demographic Changes in Rural Development in Eastern China — Comparative Studies

This study focuses on eastern China and samples the villages of Xinsi and Qingshan in Shandong Province to explore rural socioeconomic and demographic changes and their interaction with policies. Amid global urbanization, China’s rural population has migrated to cities on a large scale, causing challenges such as “rural emptying,” aging, environmental issues, and inadequate infrastructure. The government has implemented a number of policies aimed at modernizing agriculture, improving farmers’ welfare, and reducing the gap between urban and rural areas. The objective of the study was to assess the effectiveness of rural development policies, particularly in response to demographic and socio-economic changes, and the interaction between local policies and practices. Combining macroeconomic policy analysis, agricultural planning, land-use data, demographics, field visits and case studies, this paper compares the two models of rural development in inland and coastal areas. The study reveals that China’s rural areas are undergoing demographic shifts, with labor migration leading to “hollowing out,” while promoting diversification and compound transformation of the agricultural economy, such as the rise of non-agricultural industries and the development of rural tourism. At the policy level, China’s rural strategy has undergone an evolution, emphasizing agricultural modernization, optimization of industrial structure, and infrastructure construction, while current policy emphasizes innovation, sustainability, and inclusion, reflecting a “people-oriented” and “solving the problems of agriculture, rural areas and rural areas.” The policy effect is affected by regional differences, highlighting the configuration of China’s diverse geography, economy, and culture in the path of rural development and policy response. The cases of Xinsi and Qingshan villages show how rural communities can innovate multifunctional development models based on resource characteristics, connect with or challenge policy frameworks, and explore new avenues of development. Xinsi Village can actively promote the development of the surrounding service industry through cooperation with local enterprises, while Qingshan Village is expected to take advantage of coastal advantages to develop marine economy, rural tourism and modern agriculture, while protecting and inheriting the local culture.

Blue grabbing of the green crop fields: A development conundrum in southwest coastal Bangladesh

Blue grabbing of green crop fields for brackish water shrimp farming in southwest coastal Bangladesh has long been criticized for salinization, reduced crop production, occupational shift, migration, marginalization of farmers, increasing social conflict, income inequality, and unemployment. This socio-ecologically irresponsible farming system has been expanded to vast crop fields in southwest coastal Bangladesh over the last four decades. Business and policy people branded this export earning sector as the ‘White Gold’ industry, and they termed this massive farming shift as a ‘Blue Revolution’. However, the unplanned expansion of the low-yielding extensive shrimp farming is now neither competitive in the global shrimp market nor socio-ecologically sound. The current declining production trend, upcoming risks, reduced competitiveness, and the dismal prospect of the shrimp industry strongly call for a new land use policy supporting the reversion of shrimp farming into crop agriculture, and intensification, and modification of the existing extensive shrimp farm, where possible. This opinion article aims to draw policy attention to planning comprehensive crop zoning, enacting agricultural land protection and land use planning, and implementing integrated blue-green farming systems in southwest coastal Bangladesh.

Advanced evapotranspiration forecasting in Central Italy: Stacked MLP-RF algorithm and correlated Nystrom views with feature selection strategies

Evapotranspiration is a key parameter in hydrology, particularly in the field of water resources management. Reference evapotranspiration (ETo) stands as a crucial metric, embodying the influence of climate on water loss from both soil and plant surfaces. The accurate forecasting of future ETo values is paramount for informed decision-making in agricultural practices and water supply planning. The anticipation of evapotranspiration variations supports optimized irrigation, drought assessment, and efficient water allocation. Employing innovative algorithms, specifically the Multilayer Perceptron-Random Forest (MLP-RF) Stacked Model and the Correlated Nystrom Views (XNV), this study focuses on predicting ETo up to 60 days ahead in Agro Pontino, an area in Central Italy known for its flourishing agricultural production in the Mediterranean Europe. A Radial Basis Function (RBF) Neural Network serves as a benchmark. Taking into account firstly all available weather-climatic variables and subsequently adopting two different strategies for reducing the variables, based respectively on Principal Component Analysis and feature selection using Particle Swarm Optimization, three distinct sets of input variables were considered. The models based on the complete set of exogenous climatic variables demonstrated superior accuracy. However, even models relying on only mean temperature, maximum relative humidity, and shortwave solar radiation as inputs produced good results, also for the 60-day forecasting horizon, with Kling-Gupta Efficiency (KGE) and Mean Absolute Percentage Error (MAPE) equal to 0.98 and 8.356 %, respectively, in the case of MLP-RF Stacked Model. The latter consistently outperformed XNV and RBF across various combinations of input variables and forecasting horizons. Notably, the reduction in accuracy with extended forecasting horizons was mild, suggesting the potential for accurate results over significantly more extended horizons. These forecasting models facilitate precise irrigation scheduling, minimizing water wastage and conserving resources. This targeted approach enhances crop yields, quality, and environmental sustainability, rendering agriculture economically viable and adaptable to climate variability.

Machine Learning Techniques to Predict Rainfall of Vidarbh Region

Accurate daily rainfall forecasting is vital for proper agricultural planning and managing resources efficiently. This study evaluates the performance of various methods, with an emphasis on temperature and precipitation, such as linear regression, random forest, SVM regression, and XGBoost. Linear regression displays minimal RMSE and a flawless R-squared, whereas random forest performs effectively with low RMSE and encouraging R2 results. However, the presence of a negative R-squared indicates potential overfitting. The MAE, MSE, and RMSE statistics for SVM are competitive. The study draws attention to the unexplored Vidarbha region, which includes 11 districts, using Nagpur district as a representative instance. Additionally, future plans involve the utilization of deep learning models like ARIMA and LSTM to enhance rainfall prediction accuracy across Vidarbha. This investigation yields valuable insights into climate prediction, offering support for well-informed decision-making.

The Relationship between Climate, Agriculture and Land Cover in Matopiba, Brazil (1985–2020)

Climate change has been at the forefront of discussions in the scientific, economic, political, and public spheres. This study aims to analyze the impacts of climate change in the Matopiba region, assessing its relationship with land cover and land use, soybean crop production and yield, and ocean–atmosphere anomalies from 1985 to 2020. The analysis was conducted in four parts: (1) trends in annual and intra-annual climate changes, (2) the spatiotemporal dynamics of land cover and use, (3) the spatiotemporal dynamics of soybean production and yield, and (4) the relationship between climate change, agricultural practices, land cover and use, and ocean–atmosphere anomalies. Statistical analyses, including Mann–Kendall trend tests and Pearson correlation, were applied to understand these relationships comprehensively. The results indicate significant land cover and use changes over 35 years in Matopiba, with municipalities showing increasing soybean production and yield trends. There is a rising trend in annual and intra-annual maximum temperatures, alongside a decreasing trend in annual precipitation in the region. Intra-annual climate trends provide more specific insights for agricultural calendar planning. No correlation was found between the climate change trends and soybean production and yield in the evaluated data attributed to genetic and technological improvements in the region. The North Atlantic Ocean shows a positive correlation with soybean agricultural variables. Evidence suggests soybean production and yield growth under climate change scenarios. This study highlights soybeans’ adaptation and climate resilience in the Matopiba region, providing valuable insights for regional agricultural development and contributing to further research in environmental, water-related, social, and economic areas of global interest.

Assessment of the impacts of climate change on the construction of homogeneous climatic regions and ensemble climate projections using CMIP6 data over Pakistan

In the 21st century, the global environment faces significant threats due to climate change, ranging from floods and droughts to heat waves and other extreme weather events. In recent decades, Pakistan has experienced severe flooding and prolonged droughts, compounding challenges related to livelihoods, public health, and population displacement. The availability of precise and robust climate change information is paramount in addressing and mitigating the adverse impacts on a region's environment and its communities. The present study intends to assess the effects of climate change on the development of homogenous climate regions (HCRs) and then developed ensemble climate projections under the shared socioeconomic pathways (SSPs) using CMIP6 data. In the first step, HCRs were developed by integrating the cluster analysis approach and L-moment technique, identifying five distinct climatic regions. HCRs were developed using both observed data and ensemble baseline data which allowed us to thoroughly evaluate the ensemble data's performance. Subsequently, we extended our analysis to the future period (2015–2044) using ensemble climate data according to SSP2–4.5 and SSP5–8.5. Our findings indicate that, under both SSPs, certain stations exhibit discordant climatic trends in the future due to climate change. Ensemble climate projections were developed over each region for both SSPs for the duration of 2015–2044. A significant increase of 2–3.20 °C has been noted in the summer season in zone 3 and zone 5 (a region that has one of the world's largest glaciers) in maximum as well as minimum temperatures. Temperature has upward trend in the other region as well while precipitation has a mixed trend in all regions. The insights gleaned from this study hold significance and offer valuable guidance for decision-makers involved in water management, agricultural planning, and disaster management.

Regionalization of the Onset and Offset of the Rainy Season in Senegal Using Kohonen Self-Organizing Maps

This study explores the spatiotemporal variability of the onset, end, and duration of the rainy season in Senegal. These phenological parameters, crucial for agricultural planning in West Africa, exhibit high interannual and spatial variability linked to precipitation. The objective is to detect and spatially classify these indices across Senegal using different approaches. Daily precipitation data and ERA5 reanalyses from 1981 to 2018 were utilized. The employed method enables the detection of key dates. Subsequently, the Kohonen algorithm spatially classifies these indices on topological maps. The results indicate a meridional gradient of the onset, progressively later from the southeast to the northwest, whereas the end follows a north–south gradient. The duration varies from 45 days in the north to 150 days in the south. The use of self-organizing maps allows for classifying the onset, end, and duration of the season into four zones for the onset and end, and three zones for the duration of the season. They highlight the interannual irregularity of transitions, with both early and late years. The dynamic analysis underscores the complex influence of atmospheric circulation fields, notably emphasizing the importance of low-level monsoon flux. These findings have tangible implications for improving seasonal forecasts and agricultural activity planning in Senegal. They provide information on the onset, end, and duration classes for each specific zone, which can be valuable for planning crops adapted to each region.

WPS:A whole phenology-based spectral feature selection method for mapping winter crop from time-series images

Accurately obtaining the spatial distribution and planting patterns of crops is very important for agricultural planning and food security. At present, time-series images have been proved to be an effective tool to characterize crop seasonal growth patterns, and identifying crop information by measuring the time-series similarity between unknown classes and known crop phenology curves is also considered to be a useful way. However, the existing methods of selecting feature ignore the connection between each phenological stage of crops and the unique growth rules of the whole phenology, which makes it difficult to select time-series spectral features that are potentially important for crop mapping. In order to make up for this problem, a Whole Phenology-based Spectral Feature Selection (WPS) method was proposed. The aim was to select the time-series feature sets with great differences among winter crops from a large number of spectral features, so as to improve the mapping accuracy of winter rapeseed and winter wheat. Firstly, spectral separability between all classes is calculated. Secondly, the key phenological periods of winter crops were selected according to the importance of temporal features, and the spectral feature sets with high separability were selected according to the key phenological periods. Finally, a Time-weighted Dynamic Time Warping (TWDTW) algorithm was used to generate the winter rapeseed and winter wheat maps of two cities in the middle and lower reaches of the Yangtze River. The mapping accuracy of the two crops is more than 92%, which matches the crop planting area well. The research shows that combining the WPS method with the TWDTW mapping method has great potential to accurately map crop types based on satellite time-series images.