Observing Supraglacial Lakes Using Deep Learning and Planetscope Imagery

Supraglacial lakes (SGL)s result from melt water accumulation in topographic depressions on the surface of glaciers. SGLs primarily affect glacial dynamics through a positive feedback loop in which the albedo-lowering effect of SGLs can escalate surface melt leading to increases in lake extent and depth, amplifying the aforementioned albedo-lowering effect. The implications of accelerated glacial melt include increased sea level rise and modifications to ocean primary productivity. SGLs are critical indicators of surface melt and its downstream impacts and should be monitored efficiently. In situ observations and measurements of SGLs are time consuming, cost-prohibitive and difficult to scale. Earth observation data and machine learning enable scalable monitoring of SGLs through pattern detection and quantification of lake evolution over time [1]. This work presents a model developed by training a convolutional neural network with imagery and labels from NASA Operation IceBridge and predicting SGLs in high temporal and spatial resolution PlanetScope imagery.

1. INTRODUCTION

SGLs, also known as melt ponds, primarily occur during the ablation season. During this time, melt water accumulates in topographic depressions, where SGLs form and grow. As this occurs, SGLs develop the potential to introduce fragility by way of their lower albedo with respect to the adjacent ice mass. This may occur through the acceleration of melt processes, intensification of existing SGLs and instigation of ice flexure and fracturing, enabling mechanisms for melt water to enter and affect the interior of the ice column [1].