Estimating lake-climate responses from sparse data: An application to high elevation lakes

Although many studies demonstrate lake warming, few document trends from lakes with sparse data. Diel and seasonal variability of surface temperatures limit conventional trend analyses to datasets with frequent repeated observations. Thus, remote lakes, including many high elevation lakes, are underrepresented in trend analyses. We used a Bayesian technique to analyze sparse data that explicitly incorporated diel and seasonal variability. This approach allowed us to estimate lake warming in a region of limited knowledge: high elevation lakes (> 2100 m ASL) of the Southern Rocky Mountains, U.S.A. The analysis allowed for inclusion of lakes with few repeated measurements, and observations made before 1980 when more intensive lake monitoring began. We accumulated the largest dataset of high elevation lake temperatures analyzed to date. Data from 590 high elevation lakes in the Southern Rocky Mountains showed a 0.13°C decade−1 increase in surface temperatures and a 14% increase in seasonal degree days since 1955. This result is lower than other regional and global estimates of lake warming; however, it is similar to other high elevation lake studies. Our approach can be applied to other understudied regions, increasing our overall understanding of the effects of climate change on lakes and their temporal dynamics.