Located in the low-medium region of the Araguari River Basin, the study area, Indirect Influence Area (AII) of the Ferreira Gomes and Cachoeira Caldeirão Hydroelectric Power Plants in the State of Amapá, covers in its limits parts of three municipalities: Ferreira Gomes, Porto Grande and Tartarugalzinho. The objective of this work was to identify spatial scenarios of the IIA corresponding to the years of 2013 and 2017, modeling the dynamics of land use and cover that occurred during this time period, in order to simulate future landscape changes for the year 2030. (Landsat-8 (2013) satellite images merged with the panchromatic band and Sentinel-2 (2017). The modeling step was done in the software, of free license, Dinamica EGO through the model of LUCC (Land Use and Cover Change) that uses the method of Weights of Evidence. After validating and calibrating the model, the simulated maps in the 2017 models, when compared to the actual reference map, presented results of their satisfactory indexes of similarity. The analysis of the correlation between the maps in the model, generated maps of probabilities of transition between classes, of which the most influential in the model were: the probability of transition from Cerrado / Pasture to cultivated area, Cerrado / Pasture to urban area, Forest to cultivated area, Forest for Urban area and Forest for Cerrado / Grassland. From the use of Artificial Intelligence (AI) to simulate future scenarios until the year 2030, it was possible to analyze that the cultivated area class presents growth trends near the northeastern region of the study area in the area of Tartarugalzinho, right side of the Araguari river near the municipality of Porto Grande in the vicinity of BR-210; the cerrado / pasture class tends to grow even more in its predominant region on the left side of the Araguari River, near the BR-165 and also in the region near the municipal headquarters of Ferreira Gomes; the Forest area tends to lose ground due to the transition from Forest to Cerrado / Pasture and also reduces near to urban centers that tend to expand in their surroundings due to the influence of the hydroelectric projects and the growth of the municipal headquarters of Porto Grande and Ferreira gomes. It is concluded that the methodology proposed in the modeling process of the landscape dynamics in Dinamica EGO to be used for the construction of probabilistic scenarios, can help in the decision making in land use and land use planning, making AI an important technological resource for complex spatial analyzes.
|
