Abstract: Irrigation plays a critical role in paddy rice cultivation. Water availability becomes a limiting factor when irrigation do not consider crop water requirements and the predominant edaphic properties. This study was conducted in five paddy rice fields cultivation subjected to different irrigation schedules in the UsoCoello irrigation district, Tolima region, Colombia. The aim of this research was to assess the effectiveness of geophysical data, specifically apparent electrical conductivity (ECa) collected by means of the EM38-MK2 device and multispectral data based on the Normalized Difference Water Index (NDWI) from Sentinel-2A to delineate site- specific management zones (MZs). Prior to analysis, the collected variables underwent exploratory data analysis to identify and remove outliers and inliers. Further, geostatistical techniques were employed to describe the spatial variability of ECa measurements. To reduce dimensionality and account for spatial correlations in the data, a modified version of principal component analysis (PCA) known as MULTISPATI-PCA was applied. Fuzzy k-means was performed to categorize the MZ based on two sPC, where the optimal number of MZs was determined based on summary indices. Finally, each MZ was validated using Mixed Linear Models (MLM) considering spatial structures with random stratified sampling of rice yield and soil water content from TDR measurements. NDWI and ECa were poorly positively correlated (6% to 31% correlation) for most of the Plots to slightly higher negative correlation (3% to 46%) being statistically significant (Pr <0.05), while the multispectral bands showed strong correlation with the vegetation index, they presented the opposite effect (weak correlations) between ECa. Data pre-processing removed 9% (Plot 2) to 21% (Plot 1) of the total observations, lowering the coefficient of variation (CV). The spatial structure was analyzed by semivariograms, and fitted by spherical, gaussian and exponential models showing a high cross validation coefficient (CVC) of 91%, 97.5%, 76%, 70% and 90% for Plot 1 to Plot 5, respectively. Two MZs were selected and their means exhibited statistically significant (Pr<0.05) differences through Tukey’s test, providing further evidence of the effectiveness of integrating geophysical and multispectral data in delineating these zones. This methodology will be used as input for optimizing paddy rice cultivation by means of irrigation scheduling in the UsoCoello irrigation district, Tolima, Colombia.
Citación recomendada (normas APA)
Carlos Manuel Ramírez Gómez, "Integrating Geophysical and Multispectral Data to Delineate Homogeneous Management Zones in the UsoCoello Irrigation District, Colombia", Colombia:-, 2023. Consultado en línea en la Biblioteca Digital de Bogotá (https://www.bibliotecadigitaldebogota.gov.co/resources/3711954/), el día 2025-05-20.
