Geoestatística aplicada ao monitoramento de água superficial de no lago de Serpa localizado em uma área de proteção ambiental no municipio de Itacoatiara

Abstract

Water is a fundamental resource, whose essentiality is indisputable for sustaining life on our planet. In the Amazon region, the relevance of surface waters stands out, since water quality plays a crucial role in activities such as fishing and supplying water for human consumption. However, increasing pollution poses a significant challenge in this scenario. In order to face this challenge and better understand water quality, this project aims to develop an environmental monitoring tool to analyze water quality in Lake Serpa, located in the municipality of Itacoatiara/AM. Samples were collected at several points, recording geographic coordinates with the mobile topographer application to ensure a comprehensive assessment. Measurements of the physical-chemical variables temperature, pH, dissolved oxygen, turbidity, apparent color and true color were also carried out. The data obtained will be used in spatial interpolation using the ordinary kriging method using theoretical models (linear, spherical, Gaussian and exponential). To choose the best model, the cross-validation technique will be applied and the statistical indices will be calculated: mean absolute error (MAE), root mean square error (REQM), correlation coefficient (r), agreement index (d) and index trustworthy(c). The spatial correlation of each variable can be observed on the thematic maps, allowing visualization of its variation and recognizing the importance of water quality, both for the population that depends on it, directly or indirectly, and for the preservation of the environment and local biodiversity. , which allowed the analysis of current water quality standards at the site and the management of the water body under study. The analysis highlighted the varying effectiveness of different interpolation models for physicochemical variables. The analysis showed that temperature and dissolved oxygen are best interpolated using the spherical model, pH is best suited to linear interpolation, and turbidity fits best to the Gaussian model. With regard to apparent and true color, more detailed statistical and geostatistical analyzes should be applied to these variables to identify the causes of high variability although pH and dissolved oxygen do not comply with regulatory standards due to local environmental factors, such as soil erosion and decomposition of organic matter, other variables, such as apparent and true color, require more detailed analysis. Ordinary kriging was effective in visualizing the spatial distribution of variables, providing crucial information to implement protection measures for Serpa lake.