Relações entre variáveis limnológicas e reflectância do reservatório Passo Real - RS

Abstract

Despite the economic and social relevance of the reservoirs, its construction changes the water quality. In this sense, remote sensing has contributed to monitor the water quality providing spatial and temporal analysis information. The objective was to analyze the relation between limnological variables and spectral data of the Passo Real reservoir RS, Brazil, combining in situ and RapidEye data. The limnological data were collected at April 17and 18, 2009, concomitant to RapidEye acquisition (April 17, 2009). The following limnological data were determined: water transparency (Secchi disk); temperature; total of suspended solids (TSS); chlorophyll a (Chl a) and dissolved organic carbon (DOC). After atmospheric correction, for each filed sampling point was applied a 3x3 window to extract reflectance data. By using descriptive statistic, regression analysis and spatial estimative, the RapidEye reflectance and band ratio data were related to the optically active elements of water. Based on the Chl a concentration, the reservoir presented the following trophic environments: ultraoligotrophic, oligotrophic and mesotrophic. According to the CONAMA nº 357/2005 resolution, in April 2009, the Passo Real reservoir s water meet the type I. The empirical models allowed to estimative the TSS concentration e water transparency (R² = 0.87 e R² = 0.79) using the near infrared and red bands, respectively. Because of the low correlation between Chl a and the spectral data, there was a low consistency between the estimated data and field collected data (interpolated). It was found that is possible to develop empirical models to estimate the TSS concentration and water transparency using RapidEye satellite scene. The bands ratios presented similar results than the obtained from the correlation between the bands and limnological variables. The highest correlation coefficients for the TSS were obtained through the B5/B1 and B5/B2 rations (r = 0.94 e r = 0.93, respectively). The ratio B1/B3 presented the highest correlation with the water transparency (r = 0.94). The results obtained with RapidEye image, highlight the perspective to use satellite data as analysis tool for water quality monitoring.