ABSTRACT
This study aimed to estimate limnological parameters using remote sensing data. Data were sampled in 20 sites in the Orós reservoir in Ceará State, Brazil in March/2011 and August/2012. Secchi transparency, turbidity, suspended inorganic sediments (SIS), transparency, electrical conductivity of water (CE), and radiometric data were analysed. After sampling the data were submitted to correlation and regression analysis between the bidirectional reflectance factor (FR) and SIS, CE, turbidity and transparency data. For estimating SIS the adjusted model was the potential model (SIS = 860.1679*FRλ7201/0.6427) with a coefficient of determination (R2) of 0.90. For turbidity the best model was represented by euqation Turb = (FRλ720 – 0.0217)/0.0017 providing a R2 of 0.90 and the best model for transparency was Transp = (-FRλ653 + 0.171)/0.1375 with an R2 of 0.92. For the electrical conductivity of water the CE = (-FRλ632 + 0.5352)/1.6278 was the best with a R2 of 0.93. For all these models, p-value was smaller than 0.001. These models seems to be reliable and indicate that these limnological variables can be estimated using in situ remote sensing data.
Key words:
semi-arid region; monitoring of water quality; hyperspectral sensor