Criteria for the onset and breakup of summer lake stratification based on routine temperature measurements

The increase in strength and duration of summer stratification in temperate lakes is a well-described sentinel of climatic change reported in numerous lake studies. The definitions of the onset of thermal stratification in early spring and its breakup during the autumn overturn are usually binary thresholds chosen on a subjective basis. However, the transition phases between mixing and stagnation (start of summer stratification) and between stagnation and mixing (end of summer stratification) are continuous processes. Therefore, the time limits of the stratified period are definition-dependent, which complicates comparison of results from different studies. An experimental determination of the duration of summer stratification needs unified criteria to decide when stratification is established and when it terminates. We demonstrate that the termination of the stratified period (start of autumn mixing) can be detected with sufficient accuracy from near bottom temperature records at the deepest point of the lake. Nevertheless, it takes days after this moment until the whole water column is mixed completely and a stability threshold is required for the end of the stratified period. To define the beginning of the stratified period, thresholds for different stability indices were suggested previously. Although the threshold choice influences the estimated duration of stratification each year, it appears not to significantly affect the long-term trends of stratification. More important than the selection of stability thresholds is therefore the quality of observational data used for analysis. Using measured data from two dimictic Lakes in Northern Germany we show that the estimation of the stratification length depends strongly on the temporal resolution of the measurements. For comparative studies which aim to verify a shift in future temperate lake stratification experimentally, we suggest (i) to use data from high-resolution water temperature probes (discarding completely data sets with weekly, biweekly or sparser sampling period), (ii) to use a consistent stability threshold (e.g. a fixed value of Schmidt stability) to derive the duration of the stratified period uniformly, and (iii) to apply standardized algorithms in large intercomparison studies to determine indices of stratification.