Landsat Views of Sudbury (Canada) Area Acidic and Nonacidic Lakes

Publisher Website
Google Scholar
Abstract
Multivariate statistical procedures are used to establish empirical associations between acidity, visual lake water colour, dissolved organic carbon (DOC) concentrations, and Landsat-5 Thematic Mapper (TM) radiance data. Acidic lakes in an area northeast of Sudbury (Canada) are characterized by their clear, blue colours and very low DOC. With a subjective, three-class water colour grouping, 92% of the study lakes were correctly classified using TM data. Further, it is shown that DOC, the major component of water colour in this area, can be predicted within 1 mg/L of observed concentrations using TM data (multiple r = 0.93, P < 0.01). By deriving interrelationships between pH levels, water colour, and DOC, Landsat data provide a means to discriminate and map the acidic and nonacidic lakes of the study area. Examination of the reflectance characteristics of a single acidic lake (Bowland Lake) that has undergone neutralization suggests that Landsat data may be used to detect optical changes over time. However, the capability for monitoring the temporal dimension of lake acidification using satellite data has yet to be established.