Digitizing lake bathymetric data using ImageJ

Lake morphometry is a driver of limnological processes, yet digitized bathymetry is lacking for most lakes. Here, we describe a method for efficiently extracting hypsography from bathymetric maps using ImageJ. To validate our method, we compared results generated from two independent users to those obtained from digital elevation models for 100 lakes. The mean absolute difference between hypsographic curves extracted using ImageJ vs. digital elevation models (DEMs) was 0.049 (95% CI 0.041-0.056) proportion of lake area, suggesting that ImageJ provides accurate hypsography. We calculated the mean absolute difference between the two users (0.016; 95% CI: 0.011-0.021), which suggests high interobserver reliability. Finally, we compared DEMs to an interpolated hypsography using only the maximum lake depth and found large differences. We apply this method to extract data for 1012 lakes. Our data and approach will be useful where bathymetric maps exist but are not digitized.

Automated summary

This study describes an efficient method using ImageJ to extract hypsography from bathymetric maps and validates it against digital elevation models. The results show a mean absolute difference of 0.049 between hypsographic curves extracted using ImageJ and digital elevation models, indicating accurate results. The high interobserver reliability is supported by a mean absolute difference of 0.016 between two independent users. Comparisons between digital elevation models and interpolated hypsography using maximum lake depth show significant differences. The method has been successfully applied to extract data for 1012 lakes, making it valuable for digitizing bathymetric maps.