Studying Ice Phenology Changes in Rapidly Changing Glacial Lakes of Nepalese Himalayas using an Adaptive Threshold-Based Approach – Supporting Dataset

This repository hosts all research materials underlying our study on Himalayan glacial-lake ice phenology (LIP). It includes (1) Google Earth Engine scripts for training a random forest modle for delineating lake-boundary, and quantify ice fraction, and derive respective datasets (.js); (2) python scripts for model-optimization and phenology extraction (.ipynb); (2) processed high quality ice-fraction time-series (.csv) for five lakes; (3) cleaned daily time series of lake reflectance from 2000 to 2024 using MODIS data (.csv) (3) detailed phenological dates of all 4 LIP events from 2000 to 2024 and trend analysis (.xlsx) (4) detailed model outputs of the top 20 models fitting daily modis reflectance to high quality ice fraction data (.csv). Files are openly licensed (CC-BY 4.0). Users can reproduce analyses or adapt the code to new lakes by following the step-by-step notebook or take examples from associated data. Concerns can be emailed to [e-mail].