Adjusting chlorophyll-a estimates through temporal weighting based on the seasonal development of phytobiomass

Abstract

North-temperate lakes have been shown to progress through a general cycle of chlorophyll-a phenology. Because estimates of seasonal means are often based on only a few samples collected against this variable background, these estimates can be biased or uncertain. Our goal was to reduce the seasonal uncertainty and thereby produce more accurate estimates of chlorophyll concentration by defining a correction for phenological development. Time-series data from 149 lake-years were used to develop equations from which chlorophyll values could be “corrected” to the seasonal mean in relation to their particular date of measurement. However, we found the seasonal correction to be ineffectual in reducing uncertainty about nutrient-response regressions. After reviewing a number of hypotheses, we conclude that the correction derived from the average response for many lakes will be inadequate to adjust for the seasonal pattern occurring within any particular lake. This occurs because the temporal weighting correction, generated through repeated averaging, underestimates the seasonal variability which exists among individual lakes. An effective correction, if it is developed will have to be based on patterns within single lakes or possibly different lakes within a single region.