An Operational Methodology for Validating Satellite-Based Snow Albedo Measurements Using a UAV

Snow albedo is highly variable over multiple temporal and spatial scales. This variability is more pronounced in areas that experience seasonal snowpack. Satellite retrievals, physically based models and parameterizations for snow albedo all require ground-based measurements for calibration, initialization, and validation. Ground measurements are generally made using upward and downward-facing pyranometers at opportunistically located weather stations that are sparsely distributed, particularly in mountainous regions. These station-based measurements cannot capture the spatial variability of albedo across the land surface. Uncrewed Aerial Vehicles (UAVs) equipped with upward and downward-facing pyranometers provide near-surface measurements of broadband albedo that are spatially distributed across landscapes, offering improvements over in-situ sensors. At the hillslope to watershed scale albedo measurements from UAVs taken over heterogeneous terrain are a function of the spatial variability in albedo and topography within the downward-facing sensor’s field-of-view (FOV). In this research we propose methods for topographic correction of UAV snow albedo measurements and comparison to gridded satellite albedo products. These methods account for the variability of surface topography and albedo within the sensor FOV, sensor tilt, and the angular response of pyranometers. We applied the proposed methodologies to UAV snow albedo measurements collected over an alpine meadow in southwest Montana, United States (45.23°, −111.28°). Sensitivity analyses were conducted to determine the effect of altering the processing FOV (PFOV) for both topographic corrections and comparison to coincident Landsat 8-derived albedo measurements. Validation from ground-based albedo measurements showed the topographic correction to reduce albedo measurement error considerably over mildly sloping terrain. Our sensitivity analyses demonstrated that outcomes from the topographic correction and satellite comparison are highly dependent on the specified PFOV. Based on field observations and analyses of UAV albedo measurements made at different altitudes, we provide guidelines for strategizing future UAV albedo surveys. This research presents considerable advances in the standardization of UAV-based albedo measurement. We establish the foundation for future research to utilize this platform to collect near-surface validation measurements over heterogeneous terrain with high accuracy and consistency.

Latent Linear Adjustment Autoencoder v1.0: a novel method for estimating and emulating dynamic precipitation at high resolution

A key challenge in climate science is to quantify the forced response in impact-relevant variables such as precipitation against the background of internal variability, both in models and observations. Dynamical adjustment techniques aim to remove unforced variability from a target variable by identifying patterns associated with circulation, thus effectively acting as a filter for dynamically induced variability. The forced contributions are interpreted as the variation that is unexplained by circulation. However, dynamical adjustment of precipitation at local scales remains challenging because of large natural variability and the complex, nonlinear relationship between precipitation and circulation particularly in heterogeneous terrain. Building on variational autoencoders, we introduce a novel statistical model – the Latent Linear Adjustment Autoencoder (LLAAE) – that enables estimation of the contribution of a coarse-scale atmospheric circulation proxy to daily precipitation at high resolution and in a spatially coherent manner. To predict circulation-induced precipitation, the Latent Linear Adjustment Autoencoder combines a linear component, which models the relationship between circulation and the latent space of an autoencoder, with the autoencoder's nonlinear decoder. The combination is achieved by imposing an additional penalty in the cost function that encourages linearity between the circulation field and the autoencoder's latent space, hence leveraging robustness advantages of linear models as well as the flexibility of deep neural networks. We show that our model predicts realistic daily winter precipitation fields at high resolution based on a 50-member ensemble of the Canadian Regional Climate Model at 12 km resolution over Europe, capturing, for instance, key orographic features and geographical gradients. Using the Latent Linear Adjustment Autoencoder to remove the dynamic component of precipitation variability, forced thermodynamic components are expected to remain in the residual, which enables the uncovering of forced precipitation patterns of change from just a few ensemble members. We extend this to quantify the forced pattern of change conditional on specific circulation regimes. Future applications could include, for instance, weather generators emulating climate model simulations of regional precipitation, detection and attribution at subcontinental scales, or statistical downscaling and transfer learning between models and observations to exploit the typically much larger sample size in models compared to observations.

Comparison of Aerial Thermal Infrared Imagery and Helicopter Surveys of Bison (Bison bison) in Grand Canyon National Park, USA

Aerial thermal infrared (TIR) surveys are an attractive option for estimating abundances of large mammals inhabiting extensive and heterogeneous terrain. Compared to standard helicopter or fixed-wing aerial surveys, TIR flights can be conducted at higher altitudes translating into greater spatial coverage and increased observer safety; however, monetary costs are much greater. Further, there is no consensus on whether TIR surveys offer improved detection. Consequently, we performed a study to compare results of a TIR and helicopter survey of bison (Bison bison) on the Powell Plateau in Grand Canyon National Park, USA. We also compared results of both surveys to estimates obtained using a larger dataset of bison helicopter detections along the entire North Rim of the Grand Canyon. Observers in the TIR survey counted fewer individual bison than helicopter observers (101 to 127) and the TIR survey cost was 367% higher. Additionally, the TIR estimate was 18.8% lower than the estimate obtained using a larger dataset, while the comparative helicopter survey was 9.3% lower. Despite our small sample size, we found that helicopter surveys are currently the best method for estimating bison abundances in dense canopy cover sites due to ostensibly more accurate estimates and lower cost compared to TIR surveys. Additional research will be needed to evaluate the efficacy of these methods, as well as very high resolution satellite imagery, for bison populations in more open landscapes.

Ground-based lidar observations of vertical aersosol and water vapor profiles within the boundary layer over heterogeneous terrain

<p>Connecting the earth's surface with the free troposphere, the planetary boundary layer (PBL) comprises complex dynamics of turbulent behavior. This especially applies for areas with heterogeneous terrain. Relevant near-ground processes such as released energy fluxes and the emission of aerosols and trace gases directly interact with the atmosphere. Therefore, the PBL's physical state is determined both by the near-ground processes as well as entrainment of air parcels from higher layers. The mainly turbulence-driven transport of particles and properties throughout the PBL constrain a comprehensive understanding of the PBL's behavior. Hence, the energy balance closure problem as well as errors in precipitation forecast in long-term numerical weather predictions, amongst others, remain unresolved challenges. Here, ground-based lidar profiling is a well suitable method for observing the PBL, as data sampling allows for high temporal and vertical resolutions (Here: Sampling rate of 100\,Hz and 7.5\,m). During the CHEESEHEAD campaign, carried out in summer 2019, our newly developed ATMONSYS lidar performed measurements over complex terrain in northern Wisconsin. There, our lidar system was embedded in a dense network of multiple in-situ and remote sensing instruments. The central aim of this campaign was to further contribute to solve the energy balance closure problem. With the ATMONSYS lidar, vertical columns of aerosol backscatter coefficients, water vapor and temperature have been recorded. The presented work shows what the data is suitable for in terms of resolution and temporal extent in the first place. As a second point, focus is given on structure and variability of aerosol backscatter coefficient distributions and water vapor concentrations as well as their implications on the prevailing state of the PBL. Based on the presented findings, we discuss the potential and suitability of this experimental data for deriving transport processes within the PBL.</p>

Estimating Surface Melt on the Larsen Ice Shelf Using a Deep Neural Network: Opportunities and Challenges

<p>Presently, surface melt over Antarctica is estimated using climate modeling or remote sensing. However, accurately estimating surface melt remains challenging. Both climate modeling and remote sensing have limitations, particularly in the most crucial areas with intense surface melt.  The motivation of our study is to investigate the opportunities and challenges in improving the accuracy of surface melt estimation using a deep neural network. The trained deep neural network uses meteorological observations from automatic weather stations (AWS) and surface albedo observations from satellite imagery to improve surface melt simulations from the regional atmospheric climate model version 2.3p2 (RACMO2). Based on observations from three AWS at the Larsen B and C Ice Shelves, cross-validation shows a high accuracy (root mean square error = 0.898 mm.w.e.d<sup>−1</sup>, mean absolute error = 0.429 mm.w.e.d<sup>−1</sup>, and coefficient of determination = 0.958). The deep neural network also outperforms conventional machine learning models (e.g., random forest regression, XGBoost) and a shallow neural network. To compute surface melt for the entire Larsen Ice Shelf, the deep neural network is applied to RACMO2 simulations. The resulting, corrected surface melt shows a better correlation with the AWS observations in AWS 14 and 17, but not in AWS 18. Also, the spatial pattern of the surface melt is improved compared to the original RACMO2 simulation. A possible explanation for the mismatch at AWS 18 is its complex geophysical setting. Even though our study shows an opportunity to improve surface melt simulations using a deep neural network, further study is needed to refine the method, especially for complicated, heterogeneous terrain.</p>

Latent Linear Adjustment Autoencoders v1.0: A novel method for estimating and emulating dynamic precipitation at high resolution

A key challenge in climate science is to quantify the forced response in impact-relevant variables such as precipitation against the background of internal variability, both in models and observations. Dynamical adjustment techniques aim to remove unforced variability from a target variable by identifying patterns associated with circulation, thus effectively acting as a filter for dynamically-induced variability. The forced contributions are interpreted as the variation that is unexplained by circulation. However, dynamical adjustment of precipitation at local scales remains challenging because of large natural variability and the complex, nonlinear relationship between precipitation and circulation particularly in heterogeneous terrain. Building on variational autoencoders, we introduce a novel statistical model – the Latent Linear Adjustment Autoencoders v1.0 – that enables estimation of the contribution of a coarse-scale atmospheric circulation proxy to daily precipitation at high-resolution and in a spatially coherent manner. To predict circulation-induced precipitation, the Latent Linear Adjustment Autoencoders v1.0 combines a linear component, which models the relationship between circulation and the latent space of an autocoder, with the autoencoder's nonlinear decoder. The combination is achieved by imposing an additional penalty in the cost function that encourages linearity between the circulation field and the autoencoder's latent space, hence leveraging robustness advantages of linear models as well as the flexibility of deep neural networks. We show that our model predicts realistic daily winter precipitation fields at high resolution based on a 50-member ensemble of the Canadian Regional Climate Model at 12-km resolution over Europe, capturing for instance key orographic features and geographical gradients. Using the Latent Linear Adjustment Autoencoders v1.0 to remove the dynamic component of precipitation variability, forced thermodynamic components are expected to remains in the residual, which enables the uncovering of forced precipitation patterns of change from just a few ensemble members. We extend this to quantify the forced pattern of change conditional on specific circulation regimes. In addition, we briefly illustrate one of multiple possible further applications of the method: a weather generator that emulates climate model simulations of regional precipitation at high resolution by bootstrapping circulation patterns. Other potential applications include addressing detection&attribution at sub-continental scales, statistical downscaling and transfer learning between models and observations to exploit the typically much larger sample size in models compared to observations.