Título: Estimating the Water Level and Bathymetry of Lake Yahuarcocha, Ecuador using ICESat-2/ATL13 satellite laser altimetry, System Dynamics Model, and Machine Learning

Autor(es): GARRIDO SANCHEZ JOSE FERNANDO, GRANDA GUDIÑO PEDRO DAVID

Fecha de publicación: 14-jul-2023

Resumen: Monitoring the lakes and reservoirs is crucial in managing water resources due to climate change and human interventions, where available data sets on water levels and volumes are scarce. This study aims of predicting the water level height and bathymetry of lake Yahuarcocha, Imbabura, Ecuador with the use of the NASA Ice, Cloud, and Land Elevation Satellite-2 (ICESat-2) ATL13/ATL13QL since they provide observations that are suitable for water level estimation, also combining the study with System Dynamics Modelling (SD), and machine learning technique based on the extraction of points from the laser altimetry footprint at the limits of the lake vector, according to the absolute deviation method used to eliminate the outliers of the lake elevation along the orbit, and thus obtain the average values of the water level of the lake. Ya-huarache has been seriously affected in recent years, and the anthropic activities inherent to the wetlands have been the main cause of the effects, including land use, and other watershed activities influencing surface runoff and ground-waters. Therefore, predictions of lake level are crucial for their sustainable management. In this study, the SD model simulation with the PySD library is used to have the estimation of the level of precipitation, evapotranspiration, and volume of water the lake Yahuarcocha until 2030. The results indicate that the water level estimated from ATL13 is the same as the water level trend measured in situ, with an average estimation error of 30 cm, indicating that the ICE-Sat-2 ATL13 laser altimetry data has high precision, thus the approach of the proposed reaches more than 95.57%.

Palabras clave: ICESat-2 ATL13, bathymetry, water level estimation, System Dynamics Modelling, Machine Learning

DOI: https://doi.org/10.1007/978-3-031-35641-4_7

ISSN: 1865-0929

Tipo publicación: Artículo