scholarly journals Using machine learning for real-time estimates of snow water equivalent in the watersheds of Afghanistan

2018 ◽  
Vol 12 (5) ◽  
pp. 1579-1594 ◽  
Author(s):  
Edward H. Bair ◽  
Andre Abreu Calfa ◽  
Karl Rittger ◽  
Jeff Dozier
Keyword(s):  
Machine Learning ◽  
Sierra Nevada ◽  
Snow Water Equivalent ◽  
Remotely Sensed ◽  
Machine Learning Techniques ◽  
Microwave Brightness Temperature ◽  
Time Estimates ◽  
Covered Area ◽  
Spatially Distributed ◽  
Snow Water

Abstract. In the mountains, snowmelt often provides most of the runoff. Operational estimates use imagery from optical and passive microwave sensors, but each has its limitations. An accurate approach, which we validate in Afghanistan and the Sierra Nevada USA, reconstructs spatially distributed snow water equivalent (SWE) by calculating snowmelt backward from a remotely sensed date of disappearance. However, reconstructed SWE estimates are available only retrospectively; they do not provide a forecast. To estimate SWE throughout the snowmelt season, we consider physiographic and remotely sensed information as predictors and reconstructed SWE as the target. The period of analysis matches the AMSR-E radiometer's lifetime from 2003 to 2011, for the months of April through June. The spatial resolution of the predictions is 3.125 km, to match the resolution of a microwave brightness temperature product. Two machine learning techniques – bagged regression trees and feed-forward neural networks – produced similar mean results, with 0–14 % bias and 46–48 mm RMSE on average. Nash–Sutcliffe efficiencies averaged 0.68 for all years. Daily SWE climatology and fractional snow-covered area are the most important predictors. We conclude that these methods can accurately estimate SWE during the snow season in remote mountains, and thereby provide an independent estimate to forecast runoff and validate other methods to assess the snow resource.

scholarly journals Using machine learning for real-time estimates of snow water equivalent in the watersheds of Afghanistan

2017 ◽  
Author(s):  
Edward H. Bair ◽  
Andre Abreu Calfa ◽  
Karl Rittger ◽  
Jeff Dozier
Keyword(s):  
Machine Learning ◽  
Snow Water Equivalent ◽  
Remotely Sensed ◽  
Machine Learning Techniques ◽  
Microwave Brightness Temperature ◽  
Time Estimates ◽  
Covered Area ◽  
Spatially Distributed ◽  
Snow Water ◽  
Feed Forward Neural Networks

Abstract. In many mountains, snowmelt provides most of the runoff. In Afghanistan, few ground-based measurements of the snow resource exist. Operational estimates use imagery from optical and passive microwave sensors, but with their limitations. An accurate approach reconstructs spatially distributed snow water equivalent (SWE) by calculating snowmelt backward from a remotely sensed date of disappearance, but reconstructed SWE estimates are available only retrospectively; they do not provide a forecast. To estimate SWE early in the snowmelt season, we consider physiographic and remotely-sensed information as predictors and reconstructed SWE as the target. The period of analysis matches the AMSR-E radiometer's lifetime from 2003 to 2011, for the months of April through June. The spatial resolution of the predictions is 3.125 km, to match the resolution of a microwave brightness temperature product. Two machine learning techniques – bagged regression trees and feed-forward neural networks – produced similar mean results, with 0–14 % bias and 46–48 mm RMSE on average. Daily SWE climatology and fractional snow-covered area are the most important predictors. We conclude that the methods can accurately estimate SWE during the snow season in remote mountains.

scholarly journals Application of machine learning techniques for regional bias correction of snow water equivalent estimates in Ontario, Canada

2020 ◽  
Vol 24 (10) ◽  
pp. 4887-4902
Author(s):  
Fraser King ◽  
Andre R. Erler ◽  
Steven K. Frey ◽  
Christopher G. Fletcher
Keyword(s):  
Machine Learning ◽  
Linear Regression ◽  
Water Resource Management ◽  
Snow Water Equivalent ◽  
Bias Reduction ◽  
Machine Learning Techniques ◽  
Spatial Coverage ◽  
Snow Water ◽  
In Situ Observations

Abstract. Snow is a critical contributor to Ontario's water-energy budget, with impacts on water resource management and flood forecasting. Snow water equivalent (SWE) describes the amount of water stored in a snowpack and is important in deriving estimates of snowmelt. However, only a limited number of sparsely distributed snow survey sites (n=383) exist throughout Ontario. The SNOw Data Assimilation System (SNODAS) is a daily, 1 km gridded SWE product that provides uniform spatial coverage across this region; however, we show here that SWE estimates from SNODAS display a strong positive mean bias of 50 % (16 mm SWE) when compared to in situ observations from 2011 to 2018. This study evaluates multiple statistical techniques of varying complexity, including simple subtraction, linear regression and machine learning methods to bias-correct SNODAS SWE estimates using absolute mean bias and RMSE as evaluation criteria. Results show that the random forest (RF) algorithm is most effective at reducing bias in SNODAS SWE, with an absolute mean bias of 0.2 mm and RMSE of 3.64 mm when compared with in situ observations. Other methods, such as mean bias subtraction and linear regression, are somewhat effective at bias reduction; however, only the RF method captures the nonlinearity in the bias and its interannual variability. Applying the RF model to the full spatio-temporal domain shows that the SWE bias is largest before 2015, during the spring melt period, north of 44.5∘ N and east (downwind) of the Great Lakes. As an independent validation, we also compare estimated snowmelt volumes with observed hydrographs and demonstrate that uncorrected SNODAS SWE is associated with unrealistically large volumes at the time of the spring freshet, while bias-corrected SWE values are highly consistent with observed discharge volumes.

scholarly journals A Landsat-Era Sierra Nevada Snow Reanalysis (1985–2015)

2016 ◽  
Vol 17 (4) ◽  
pp. 1203-1221 ◽  
Author(s):  
Steven A. Margulis ◽  
Gonzalo Cortés ◽  
Manuela Girotto ◽  
Michael Durand
Keyword(s):  
Sierra Nevada ◽  
Snow Water Equivalent ◽  
Reanalysis Dataset ◽  
In Situ Data ◽  
Snow Covered Area ◽  
Covered Area ◽  
Snow Water ◽  
Fully Bayesian ◽  
Mean Square Errors

Abstract A newly developed state-of-the-art snow water equivalent (SWE) reanalysis dataset over the Sierra Nevada (United States) based on the assimilation of remotely sensed fractional snow-covered area data over the Landsat 5–8 record (1985–2015) is presented. The method (fully Bayesian), resolution (daily and 90 m), temporal extent (31 years), and accuracy provide a unique dataset for investigating snow processes. The verified dataset (based on a comparison with over 9000 station years of in situ data) exhibited mean and root-mean-square errors less than 3 and 13 cm, respectively, and correlation greater than 0.95 compared with in situ SWE observations. The reanalysis dataset was used to characterize the peak SWE climatology to provide a basic accounting of the stored snowpack water in the Sierra Nevada over the last 31 years. The pixel-wise peak SWE volume over the domain was found to be 20.0 km3 on average with a range of 4.0–40.6 km3. The ongoing drought in California contains the two lowest snowpack years (water years 2014 and 2015) and three of the four driest years over the examined record. It was found that the basin-average peak SWE, while underestimating the total water storage in snowpack over the year, accurately captures the interannual variability in stored snowpack water. However, the results showed that the assumption that 1 April SWE is representative of the peak SWE can lead to significant underestimation of basin-average peak SWE both on an average (21% across all basins) and on an interannual basis (up to 98% across all basin years).

scholarly journals Investigating the Variability of High-Elevation Seasonal Orographic Snowfall Enhancement and Its Drivers across Sierra Nevada, California

2018 ◽  
Vol 19 (1) ◽  
pp. 47-67 ◽  
Author(s):  
Laurie S. Huning ◽  
Steven A. Margulis
Keyword(s):  
Sierra Nevada ◽  
Snow Water Equivalent ◽  
High Elevation ◽  
Mountain Range ◽  
Distributed Data ◽  
Mountainous Regions ◽  
Average Maximum ◽  
Spatially Distributed ◽  
Orographic Enhancement ◽  
Snow Water

Abstract While orographically driven snowfall is known to be important in mountainous regions, a complete understanding of orographic enhancement from the basin to the mountain range scale is often inhibited by limited distributed data and spatial and/or temporal resolutions. A novel, 90-m spatially distributed snow water equivalent (SWE) reanalysis was used to overcome these limitations. Leveraging this SWE information, the interannual variability of orographic gradients in cumulative snowfall (CS) was investigated over 14 windward (western) basins in the Sierra Nevada in California from water years 1985 to 2015. Previous studies have not provided a detailed multidecadal climatology of orographic CS gradients or compared wet-year and dry-year orographic CS patterns, distributions, and gradients across an entire mountain range. The magnitude of seasonal CS gradients range from over 15 cm SWE per 100-m elevation to under 1 cm per 100 m with a 31-yr average of 6.1 cm per 100 m below ~2500 m in the western basins. The 31-yr average CS gradients generally decrease in higher elevation zones across the western basins and become negative at the highest elevations. On average, integrated vapor transport and zonal winds at 700 hPa are larger during wet years, leading to higher orographically driven CS gradients across the Sierra Nevada than in dry years. Below ~2500 m, wet years yield greater enhancement (relative to dry years) by factors of approximately 2 and 3 in the northwestern and southwestern basins, respectively. Overall, the western Sierra Nevada experiences about twice as much orographic enhancement during wet years as in dry years below the elevation corresponding to the 31-yr average maximum CS.

scholarly journals A Particle Batch Smoother Approach to Snow Water Equivalent Estimation

2015 ◽  
Vol 16 (4) ◽  
pp. 1752-1772 ◽  
Author(s):  
Steven A. Margulis ◽  
Manuela Girotto ◽  
Gonzalo Cortés ◽  
Michael Durand
Keyword(s):  
Data Assimilation ◽  
Sierra Nevada ◽  
Forest Cover ◽  
Snow Water Equivalent ◽  
Ground Surface ◽  
Snow Covered Area ◽  
Covered Area ◽  
Sensitivity Tests ◽  
American River ◽  
Snow Water

Abstract This paper presents a newly proposed data assimilation method for historical snow water equivalent SWE estimation using remotely sensed fractional snow-covered area fSCA. The newly proposed approach consists of a particle batch smoother (PBS), which is compared to a previously applied Kalman-based ensemble batch smoother (EnBS) approach. The methods were applied over the 27-yr Landsat 5 record at snow pillow and snow course in situ verification sites in the American River basin in the Sierra Nevada (United States). This basin is more densely vegetated and thus more challenging for SWE estimation than the previous applications of the EnBS. Both data assimilation methods provided significant improvement over the prior (modeling only) estimates, with both able to significantly reduce prior SWE biases. The prior RMSE values at the snow pillow and snow course sites were reduced by 68%–82% and 60%–68%, respectively, when applying the data assimilation methods. This result is encouraging for a basin like the American where the moderate to high forest cover will necessarily obscure more of the snow-covered ground surface than in previously examined, less-vegetated basins. The PBS generally outperformed the EnBS: for snow pillows the PBS RMSE was ~54% of that seen in the EnBS, while for snow courses the PBS RMSE was ~79% of the EnBS. Sensitivity tests show relative insensitivity for both the PBS and EnBS results to ensemble size and fSCA measurement error, but a higher sensitivity for the EnBS to the mean prior precipitation input, especially in the case where significant prior biases exist.

scholarly journals Multi-sensor analysis of monthly gridded snow precipitation on alpine glaciers

2021 ◽  
Vol 18 ◽  
pp. 7-20
Author(s):  
Rebecca Gugerli ◽  
Matteo Guidicelli ◽  
Marco Gabella ◽  
Matthias Huss ◽  
Nadine Salzmann
Keyword(s):  
Snow Water Equivalent ◽  
Cosmic Ray ◽  
Remotely Sensed ◽  
High Elevation ◽  
High Mountain ◽  
Data Set ◽  
Novel Approach ◽  
Spatially Distributed ◽  
Precipitation Estimates ◽  
Snow Water

Abstract. Accurate and reliable solid precipitation estimates for high mountain regions are crucial for many research applications. Yet, measuring snowfall at high elevation remains a major challenge. In consequence, observational coverage is typically sparse, and the validation of spatially distributed precipitation products is complicated. This study presents a novel approach using reliable daily snow water equivalent (SWE) estimates by a cosmic ray sensor on two Swiss glacier sites to assess the performance of various gridded precipitation products. The ground observations are available during two and four winter seasons. The performance of three readily-available precipitation data products based on different data sources (gauge-based, remotely-sensed, and re-analysed) is assessed in terms of their accuracy compared to the ground reference. Furthermore, we include a data set, which corresponds to the remotely-sensed product with a local adjustment to independent SWE measurements. We find a large bias of all precipitation products at a monthly and seasonal resolution, which also shows a seasonal trend. Moreover, the performance of the precipitation products largely depends on in situ wind direction during snowfall events. The varying performance of the three precipitation products can be partly explained with their compilation background and underlying data basis.

Sign in / Sign up

Export Citation Format

Share Document