MANAGEMENT SYSTEM BY SATELLITE IMAGES FOR GUAVA (PSIDIUM GUAJAVA L.) ORCHARDS ON YIELD ESTIMATES BASED ON THE CHARACTERIZATION OF GUAVA TREES BY PHENOTYPIC STAGE

Abstract

The management system which is based on handling GIS database files and remotely sensed imageries on Geographical Information System (GIS) environment is an essential tool to guide and regulate the production systems and marketing activities of product chains. In the case of guava, the management system has to be fed with a specific and structured field database and its metadata to verify the phenotypic status of guava trees and to characterize the production system applied by the producer in agronomic terms (coordinate data, planting frame, vigor, health etc). This data is useful to determine agronomic conditions to be incorporated as a field test in a supervised classification process on SPOT satellite images (PAN and MS) which is used as a mean to scale the punctual field observation data. As a numerical value, the raster structure of satellite images and the numeric value for each cell represents guava photosynthetic activity rate. Along with this, the isolation of guava's phenotypic signature in DN units will feed the supervised classification procedure allowing the user to generate thematic maps per class. On GIS environment, a multi-assessment analysis is applied in order to assign a weight to each component on the project. The analysis is based on spatial analysis vicinity functions applied to guava orchard locations, proximity to collecting centers, roads, and highways. The visual immediacy on thematic maps will allow the decision makers to implement their own criteria analysis to plan the market destination of guava production volume weeks after harvest (fresh market or industry) in Calvillo County.