BEGIN:VCALENDAR VERSION:2.0 PRODID:-//pretalx//talks.osgeo.org//foss4g-2026//speaker//F9RABK BEGIN:VTIMEZONE TZID:JST BEGIN:STANDARD DTSTART:20000101T000000 RRULE:FREQ=YEARLY;BYMONTH=1 TZNAME:JST TZOFFSETFROM:+0900 TZOFFSETTO:+0900 END:STANDARD END:VTIMEZONE BEGIN:VEVENT UID:pretalx-foss4g-2026-MFLPEV@talks.osgeo.org DTSTART;TZID=JST:20260903T113000 DTEND;TZID=JST:20260903T120000 DESCRIPTION:Mapping and monitoring agricultural crop health\, particularly for sugarcane\, is essential for improving yield quality and ensuring sust ainable crop management. Sugarcane plays a critical role in agricultural e conomies and bioenergy production in many tropical and subtropical countri es\, including Thailand. However\, crop productivity and quality are stron gly influenced by soil fertility and nutrient availability. Accurate and t imely information on soil nutrients is therefore crucial for effective cro p management and decision-making. Traditional soil analysis methods rely h eavily on labor-intensive field sampling and laboratory testing\, which ar e often time-consuming\, costly\, and spatially limited. As a result\, the se approaches are not always suitable for large-scale monitoring of soil c onditions across extensive agricultural landscapes. Recent advances in Ear th Observation (EO) technologies provide new opportunities to overcome the se limitations. Satellite-based EO now offer high spatial\, spectral\, and temporal resolution data that can be used to monitor crop health and soil conditions at regional scales. In particular\, hyperspectral satellite ha s shown strong potential for estimating soil properties because of its abi lity to capture detailed spectral signatures associated with soil composit ion\, moisture content\, and nutrient levels. Hyperspectral sensors offer continuous spectral information across hundreds of narrow bands\, enabling the detection of subtle variations in soil characteristics that may not b e captured by multispectral sensors. When combined with machine learning a lgorithms\, hyperspectral data can be used to develop predictive models fo r estimating soil nutrients and other agronomic parameters at the field le vels across large agricultural areas. Despite these technological advancem ents\, several challenges remain for implementing satellite-based soil mon itoring in Thailand. The northeastern region of the country is characteriz ed by complex topography\, heterogeneous land-use patterns\, fragmented sm allholder farms\, and highly variable weather conditions. These factors co mplicate the accurate mapping of soil properties and crop conditions using remote sensing data. In addition\, sugarcane cultivation in this region o ften occurs within relatively small and dispersed field plots\, making it difficult to capture field-level variability using conventional monitoring approaches. Therefore\, the integration of hyperspectral satellite data w ith advanced machine learning models offers a promising approach to improv e soil nutrient estimation and support precision agriculture practices in the region. This study aims to develop predictive models for estimating ke y soil nutrients in sugarcane fields across Northeast Thailand using PRISM A hyperspectral imagery and the random forest (RF) algorithm. Specifically \, the study focuses on estimating six important soil properties: soil pH\ , soil organic matter (SOM)\, electrical conductivity (EC)\, nitrogen (N)\ , phosphorus (P)\, and potassium (K). These soil parameters are critical i ndicators of soil fertility and directly influence crop growth\, nutrient uptake\, and overall yield potential. The analysis was conducted using hyp erspectral imagery (239 bands) acquired from the PRISMA satellite platform \, which provides high-resolution spectral data suitable for environmental and agricultural monitoring applications. To support model development an d validation\, field data were collected from sugarcane-growing areas acro ss the study region. A total of 46 soil sampling plots were established wi thin representative sugarcane fields. Field surveys were conducted between 1st and 25th May 2025\, during which soil samples were collected at an ap proximate depth of 20 cm. The collected soil samples were subsequently ana lyzed to determine the corresponding soil nutrient properties\, including pH\, SOM\, EC\, N\, P\, and K thought laboratory. These ground measurement s served as reference data for training and validating the predictive mode ls. In this study\, a RF was implemented to estimate soil nutrients from h yperspectral satellite data. The RF algorithm is widely recognized for its robustness\, ability to handle nonlinear relationships\, and strong perfo rmance when working with high-dimensional datasets\, like hyperspectral im agery. The model development process included feature extraction and optim ization to identify the most relevant spectral variables associated with s oil nutrient conditions. To evaluate model performance\, the dataset in th is study was divided into training and validation subsets. Specifically\, 37 soil samples (80%) were used for training the model\, while the remaini ng 9 samples (20%) were reserved for independent validation of the model r esults. \nThe experimental results demonstrated that the RF-based predicti ve models achieved varying levels of accuracy for different soil propertie s. The coefficient of determination (R²) values ranged from 0.30 to 0.80 across the estimated parameters. Higher predictive performance was general ly observed for N and OM variables\, which exhibit stronger spectral respo nses in hyperspectral imagery. In contrast\, P and EC parameters showed re latively lower predictive accuracy due to weaker spectral signatures and p otential environmental influences. Nevertheless\, the resulting spatial di stribution maps of soil fertility indicators exhibited consistent and mean ingful spatial patterns when compared with the ground datasets. The genera ted maps provide valuable insights into the spatial variability of soil fe rtility conditions across sugarcane fields in Northeast Thailand. These sp atial patterns can help farmers\, agricultural planners\, and policymakers identify areas with nutrient deficiencies or potential soil management is sues. By integrating remote sensing-based monitoring with field observatio ns\, the developed approach offers a scalable framework for large-area soi l nutrients assessment and crop monitoring. In addition\, this work develo ped a spatial recommendation framework based on the derived nutrient maps and crop condition indicators. This framework aims to support precision ag riculture practices by providing location-specific information for crop ma nagement. For example\, farmers can use these spatial recommendations to o ptimize fertilizer application\, adjust irrigation strategies\, and improv e overall crop management during different growth stages. The targeted int erventions can help improve resource-use efficiency\, reduce environmental impacts\, and enhance crop productivity. Overall\, this study demonstrate s the potential of integrating PRISMA hyperspectral satellite data with th e efficient RF model for monitoring soil nutrients at the field levels ove r the large-scale. The proposed approach provides a cost-effective and sca lable alternative to traditional soil sampling methods\, enabling continuo us monitoring of soil fertility conditions at the field level. By supporti ng more informed and timely agricultural management decisions\, this frame work contributes to the advancement of precision agriculture and sustainab le sugarcane production in Thailand. Ultimately\, the proposed workflow ca n help improve crop health\, optimize fertilizer management\, and increase sugarcane yields across diverse agricultural landscapes.\n\nKeywords: Sug arcane crop\, random forest\, PRISMA hyperspectral\, soil nutrients\, prec ision agriculture DTSTAMP:20260717T220450Z LOCATION:Cosmos1 SUMMARY:Synergistic use of PRISMA hyperspectral data and a random forest al gorithm to map soil nutrients in sugarcane fields in Northeast Thailand - Jaturong Som-ard URL:https://talks.osgeo.org/foss4g-2026/talk/MFLPEV/ END:VEVENT END:VCALENDAR