Arctic amplification has altered the climate patterns both regionally and globally, resulting in more frequent and more intense extreme weather events in the past few decades. The essential part of Arctic amplification is the unprecedented sea ice loss as demonstrated by satellite observations. Accurately forecasting Arctic sea ice from sub-seasonal to seasonal scales has been a major research question with fundamental challenges at play. In addition to physics-based Earth system models, researchers have been applying multiple statistical and machine learning models for sea ice forecasting. Looking at the potential of data-driven approaches to study sea ice variations, we propose MT-IceNet – a UNet-based spatial and multi-temporal (MT) deep learning model for forecasting Arctic sea ice concentration (SIC). The model uses an encoder-decoder architecture with skip connections and processes multi-temporal input streams to regenerate spatial maps at future timesteps. Using bi-monthly and monthly satellite retrieved sea ice data from NSIDC as well as atmospheric and oceanic variables from ERA5 reanalysis product during 1979-2021, we show that our proposed model provides promising predictive performance for per-pixel SIC forecasting with up to 60% decrease in prediction error for a lead time of 6 months as compared to its state-of-the-art counterparts.

Critical natural resources and processes in the Arctic depend heavily on sea ice. Thus, accurate and timely predictions of Arctic sea ice changes is important. Arctic sea ice forecasting involves two connected tasks: predicting sea ice concentration (SIC) at each pixel and predicting overall sea ice extent (SIE). Instead of having two separate models for these two forecasting tasks, in this paper we study how to use multi-task learning techniques and leverage the connections between ice concentration and ice extent to improve accuracy for both forecasting tasks. Because of the spatiotemporal nature of the data, we designed two novel multi-task learning models based on the CNN and ConvLSTM, respectively. Further, in conjunction with multi-task models, we developed custom loss functions which train the models to ignore land pixels and optimize for both concentration and extent when making predictions. Our experiments show that multi-task models provide better accuracy for a 1-month lead time than models that predict sea ice extent and concentration separately. Our accuracies are better than or comparable to results in related state-of-the-art studies. Our best model in SIC prediction outperformed the best existing SIC prediction model in the literature with 1.78% less error, and our best model in SIE prediction outperformed the best existing SIE prediction model with 0.283 million km2less error.

Domain adaptation techniques have been developed to handle data from multiple sources or domains. Most existing domain adaptation models assume that source and target domains are homogeneous, i.e., they have the same feature space. Nevertheless, many real world applications often deal with data from heterogeneous domains that come from completely different feature spaces. In our remote sensing application, data in source domain (from an active spaceborne Lidar sensor CALIOP onboard CALIPSO satellite) contain 25 attributes, while data in target domain (from a passive spectroradiometer sensor VIIRS onboard Suomi-NPP satellite) contain 20 different attributes. CALIOP has better representation capability and sensitivity to aerosol types and cloud phase, while VIIRS has wide swaths and better spatial coverage but has inherent weakness in differentiating atmospheric objects on different vertical levels. To address this mismatch of features across the domains/sensors, we propose a novel end-to-end deep domain adaptation with domain mapping and correlation alignment (DAMA) to align the heterogeneous source and target domains in active and passive satellite remote sensing data. It can learn domain invariant representation from source and target domains by transferring knowledge across these domains, and achieve additional performance improvement by incorporating weak label information into the model (DAMA-WL). Our experiments on a collocated CALIOP and VIIRS dataset show that DAMA and DAMA-WL can achieve higher classification accuracy in predicting cloud types.