Search for: All records

Abstract. Systematic biases and coarse resolutions are major limitations ofcurrent precipitation datasets. Many deep learning (DL)-based studies havebeen conducted for precipitation bias correction and downscaling. However,it is still challenging for the current approaches to handle complexfeatures of hourly precipitation, resulting in the incapability ofreproducing small-scale features, such as extreme events. This studydeveloped a customized DL model by incorporating customized loss functions,multitask learning and physically relevant covariates to bias correct anddownscale hourly precipitation data. We designed six scenarios tosystematically evaluate the added values of weighted loss functions,multitask learning, and atmospheric covariates compared to the regular DLand statistical approaches. The models were trained and tested using theModern-era Retrospective Analysis for Research and Applications version 2(MERRA2) reanalysis and the Stage IV radar observations over the northerncoastal region of the Gulf of Mexico on an hourly time scale. We found thatall the scenarios with weighted loss functions performed notably better thanthe other scenarios with conventional loss functions and a quantilemapping-based approach at hourly, daily, and monthly time scales as well asextremes. Multitask learning showed improved performance on capturing finefeatures of extreme events and accounting for atmospheric covariates highlyimproved model performance at hourly and aggregated time scales, while theimprovement is not as large as from weighted loss functions. We show thatthe customized DL model can better downscale and bias correct hourlyprecipitation datasets and provide improved precipitation estimates at finespatial and temporal resolutions where regular DL and statistical methodsexperience challenges. 

Spatial resolution is critical for observing and monitoring environmental phenomena. Acquiring high-resolution bathymetry data directly from satellites is not always feasible due to limitations on equipment, so spatial data scientists and researchers turn to single image super-resolution (SISR) methods that utilize deep learning techniques as an alternative method to increase pixel density. While super resolution residual networks (e.g., SR-ResNet) are promising for this purpose, several challenges still need to be addressed: (1) Earth data such as bathymetry is expensive to obtain and relatively limited in its data record amount; (2) certain domain knowledge needs to be complied with during model training; (3) certain areas of interest require more accurate measurements than other areas. To address these challenges, following the transfer learning principle, we study how to leverage an existing pre-trained super-resolution deep learning model, namely SR-ResNet, for high-resolution bathymetry data generation. We further enhance the SR-ResNet model to add corresponding loss functions based on domain knowledge. To let the model perform better for certain spatial areas, we add additional loss functions to increase the penalty of the areas of interest. Our experiments show our approaches achieve higher accuracy than most baseline models when evaluating using metrics including MSE, PSNR, and SSIM. 

Abstract. The Pleistocene sand sea on the Arctic Coastal Plain (ACP) ofnorthern Alaska is underlain by an ancient sand dune field, a geologicalfeature that affects regional lake characteristics. Many of these lakes,which cover approximately 20 % of the Pleistocene sand sea, are relativelydeep (up to 25 m). In addition to the natural importance of ACP sand sealakes for water storage, energy balance, and ecological habitat, the needfor winter water for industrial development and exploration activities makeslakes in this region a valuable resource. However, ACP sand sea lakes havereceived little prior study. Here, we collect in situ bathymetric data totest 12 model variants for predicting sand sea lake depth based on analysisof Landsat-8 Operational Land Imager (OLI) images. Lake depth gradients weremeasured at 17 lakes in midsummer 2017 using a Humminbird 798ci HD SI Comboautomatic sonar system. The field-measured data points were compared tored–green–blue (RGB) bands of a Landsat-8 OLI image acquired on 8 August2016 to select and calibrate the most accurate spectral-depth model for eachstudy lake and map bathymetry. Exponential functions using a simple bandratio (with bands selected based on lake turbidity and bed substrate)yielded the most successful model variants. For each lake, the most accuratemodel explained 81.8 % of the variation in depth, on average. Modeled lakebathymetries were integrated with remotely sensed lake surface area toquantify lake water storage volumes, which ranged from 1.056×10-3 to 57.416×10-3 km3. Due to variations in depthmaxima, substrate, and turbidity between lakes, a regional model iscurrently infeasible, rendering necessary the acquisition of additional insitu data with which to develop a regional model solution. Estimating lakewater volumes using remote sensing will facilitate better management ofexpanding development activities and serve as a baseline by which toevaluate future responses to ongoing and rapid climate change in the Arctic.All sonar depth data and modeled lake bathymetry rasters can be freelyaccessed at https://doi.org/10.18739/A2SN01440 (Simpson and Arp, 2018) andhttps://doi.org/10.18739/A2HT2GC6G (Simpson, 2019), respectively. 

Climate and weather data such as precipitation derived from Global Climate Models (GCMs) and satellite observations are essential for the global and local hydrological assessment. However, most climatic popular precipitation products (with spatial resolutions coarser than 10km) are too coarse for local impact studies and require “downscaling” to obtain higher resolutions. Traditional precipitation downscaling methods such as statistical and dynamic downscaling require an input of additional meteorological variables, and very few are applicable for downscaling hourly precipitation for higher spatial resolution. Based on dynamic dictionary learning, we propose a new downscaling method, PreciPatch, to address this challenge by producing spatially distributed higher resolution precipitation fields with only precipitation input from GCMs at hourly temporal resolution and a large geographical extent. Using aggregated Integrated Multi-satellitE Retrievals for GPM (IMERG) data, an experiment was conducted to evaluate the performance of PreciPatch, in comparison with bicubic interpolation using RainFARM—a stochastic downscaling method, and DeepSD—a Super-Resolution Convolutional Neural Network (SRCNN) based downscaling method. PreciPatch demonstrates better performance than other methods for downscaling short-duration precipitation events (used historical data from 2014 to 2017 as the training set to estimate high-resolution hourly events in 2018). 

Observations taken over the last few decades indicate that dramatic changes are occurring in the Arctic‐Boreal Zone (ABZ), which are having significant impacts on ABZ inhabitants, infrastructure, flora and fauna, and economies. While suitable for detecting overall change, the current capability is inadequate for systematic monitoring and for improving process‐based and large‐scale understanding of the integrated components of the ABZ, which includes the cryosphere, biosphere, hydrosphere, and atmosphere. Such knowledge will lead to improvements in Earth system models, enabling more accurate prediction of future changes and development of informed adaptation and mitigation strategies. In this article, we review the strengths and limitations of current space‐based observational capabilities for several important ABZ components and make recommendations for improving upon these current capabilities. We recommend an interdisciplinary and stepwise approach to develop a comprehensive ABZ Observing Network (ABZ‐ON), beginning with an initial focus on observing networks designed to gain process‐based understanding for individual ABZ components and systems that can then serve as the building blocks for a comprehensive ABZ‐ON.