Search

There has been little rigorous investigation of the transferability of existing empirical water clarity models developed at one location or time to other lakes and dates of imagery with differing conditions. Machine learning methods have not been widely adopted for analysis of lake optical properties such as water clarity, despite their successful use in many other applications of environmental remote sensing. This study compares model performance for a random forest (RF) machine learning algorithm and a simple 4-band linear model with 13 previously published empirical non-machine learning algorithms. We use Landsat surface reflectance product data aligned with spatially and temporally co-located in situ Secchi depth observations from northeastern USA lakes over a 34-year period in this analysis. To evaluate the transferability of models across space and time, we compare model fit using the complete dataset (all images and samples) to a single-date approach, in which separate models are developed for each date of Landsat imagery with more than 75 field samples. On average, the single-date models for all algorithms had lower mean absolute errors (MAE) and root mean squared errors (RMSE) than the models fit to the complete dataset. The RF model had the highest pseudo-R2 for the single-date approach as well as the complete dataset, suggesting that an RF approach outperforms traditional linear regression-based algorithms when modeling lake water clarity using satellite imagery. [ABSTRACT FROM AUTHOR]

Lakes in arid regions play an important role in regional water cycles and are a vital economic resource, but can fluctuate widely in area and volume. This study demonstrates the use of a multisensor satellite remote sensing method for the comprehensive monitoring of lake surface areas, water levels, and volume for the Toshka Lakes in southern Egypt, from lake formation in 1998 to mid-2017. Two spectral water indices were used to construct a daily time-series of surface area from the Advanced Very High Resolution Radiometer (AVHRR) and the Moderate Resolution Imaging Spectroradiometer (MODIS), validated by higher-resolution Landsat images. Water levels were obtained from analysis of digital elevation models from the Shuttle Radar Topography Mission (SRTM) and the Advanced Spaceborne Thermal Emission and Reflection Radiometer (ASTER), validated with ICESat Geoscience Laser Altimeter System (GLAS) laser altimetry. Total lake volume peaked at 26.54 × 109 m3 in December 2001, and declined to 0.76 × 109 m3 by August 2017. Evaporation accounted for approximately 86% of the loss, and groundwater recharge accounted for 14%. Without additional inflows, the last remaining lake will likely disappear between 2020 and 2022. The Enhanced Lake Index, a water index equivalent to the Enhanced Vegetation Index, was found to have lower noise levels than the Normalized Difference Lake Index. The results show that multi-platform satellite remote sensing provides an efficient method for monitoring the hydrology of lakes. [ABSTRACT FROM AUTHOR]