H33J-0967:
Is the frequency of algal blooms increasing in oligotrophic lakes in temperate forests?

Wednesday, 17 December 2014
Aleksey Paltsev and Irena F Creed, University of Western Ontario, London, ON, Canada
Abstract:
Oligotrophic lakes in the temperate forests of eastern North America appear to be experiencing an increase in the frequency and duration of algal blooms. This has been the focus of numerous public and government reports, resulting in heightened public concern for reporting of algal blooms. There is a vital need for detailed historical survey of numerous lakes, covering large spatial scales (the scale of region, province, or entire country) and temporal scales (decades) to determine if public observations are accurate. We used a remote sensing approach to: (1) develop regression models that relate Landsat imagery reflectance to chlorophyll-a (Chl-a) as a proxy of algal biomass of lakes; (2) apply these models to estimate Chl-a in lakes at the northern edge of the temperate forest biome in central Ontario over a 28 year period (1984–2011). The linear regression model was built on the basis of the normalized exoatmospheric reflectance values acquired from the utility of Landsat TM and ETM imagery and in situ measurements. Landsat band 3 (red) showed the strongest correlation with in situ data explaining 84% of the variance in Chl-a (r2 = 0.84, p <0.001). We applied this model to all lakes within the region selected from atmospherically corrected Landsat data for the peak algal bloom period (late July to early November) for the entire 28 years. A time series revealed a cyclic stationary pattern in the average Chl-a. This pattern followed the regional patterns of major droughts, especially for the first part of the time period, making climate a major driver in the formation of algal biomass in lakes that, in turn, can lead to the rise of algal blooms. However this climate driver appeared to become less predictable, with elevated algal biomass occurring in both normal and drought years, later in the record.