摘要

A nearly three-decade (May-October) survey of the Secchi disk depth (SDD) and chlorophyll-a concentration (Chl-a) was completed for Lake Diefenbaker from 1984 to 2012 with Landsat imagery. Novel predictive models were developed by comparing 2011 and 2012 field data to reflectance data from Landsat imagery. These models were applied to the archived imagery (1984-2012) to predict SDD and Chl-a. Predicted values were also tested independently with historical field data. Predicted and actual measurements for SDD and Chl-a were significantly related (R-2 = 0.84 and 0.57; RMSE 0.59 m and 1.43 mu g/L-1, respectively). These models are relevant because they can predict past and future water quality for Lake Diefenbaker without the need for reference field data. Water clarity and Chl-a concentration were positively correlated, suggesting that light limitation can occur in this reservoir. Predicted data showed that both SDD and mean Chl-a concentration have decreased in the reservoir from 1984 to 2012. Temporally, the decreases in SDD and Chl-a followed a decreasing sigmoid pattern. Mean seasonal and mean yearly flow caused significant reductions in water clarity and Chl-a concentration (R-2 = 0.28 and 0.25, p = 0.048 and 0.021, respectively). Algal blooms were observed in the reservoir during the study period, and were most frequently found in the Qu'Appelle arm of the reservoir. However, despite concerns from local residents, we did not find evidence that suggests an increase in frequency over the 29 year period.

  • 出版日期2015
  • 单位Saskatoon; Saskatchewan