Inference and uncertainty of snow depth spatial distribution at the kilometre scale in the Colorado Rocky Mountains: the effects of sample size,random sampling,predictor quality,and validation procedures

Historically, observing snow depth over large areas has been difficult. When snow depth observations are sparse, regression models can be used to infer the snow depth over a given area. Data sparsity has also left many important questions about such inference unexamined. Improved inference, or estimation, of snow depth and its spatial distribution from a given set of observations can benefit a wide range of applications from water resource management, to ecological studies, to validation of satellite estimates of snow pack. The development of Light Detection and Ranging (LiDAR) technology has provided non‐sparse snow depth measurements, which we use in this study, to address fundamental questions about snow depth inference using both sparse and non‐sparse observations. For example, when are more data needed and when are data redundant? Results apply to both traditional and manual snow depth measurements and to LiDAR observations. Through sampling experiments on high‐resolution LiDAR snow depth observations at six separate 1.17‐km² sites in the Colorado Rocky Mountains, we provide novel perspectives on a variety of issues affecting the regression estimation of snow depth from sparse observations. We measure the effects of observation count, random selection of observations, quality of predictor variables, and cross‐validation procedures using three skill metrics: percent error in total snow volume, root mean squared error (RMSE), and R². Extremes of predictor quality are used to understand the range of its effect; how do predictors downloaded from internet perform against more accurate predictors measured by LiDAR? Whereas cross validation remains the only option for validating inference from sparse observations, in our experiments, the full set of LiDAR‐measured snow depths can be considered the ‘true’ spatial distribution and used to understand cross‐validation bias at the spatial scale of inference. We model at the 30‐m resolution of readily available predictors, which is a popular spatial resolution in the literature. Three regression models are also compared, and we briefly examine how sampling design affects model skill. Results quantify the primary dependence of each skill metric on observation count that ranges over three orders of magnitude, doubling at each step from 25 up to 3200. Whereas uncertainty (resulting from random selection of observations) in percent error of true total snow volume is typically well constrained by 100–200 observations, there is considerable uncertainty in the inferred spatial distribution (R²) even at medium observation counts (200–800). We show that percent error in total snow volume is not sensitive to predictor quality, although RMSE and R² (measures of spatial distribution) often depend critically on it. Inaccuracies of downloaded predictors (most often the vegetation predictors) can easily require a quadrupling of observation count to match RMSE and R² scores obtained by LiDAR‐measured predictors. Under cross validation, the RMSE and R² skill measures are consistently biased towards poorer results than their true validations. This is primarily a result of greater variance at the spatial scales of point observations used for cross validation than at the 30‐m resolution of the model. The magnitude of this bias depends on individual site characteristics, observation count (for our experimental design), and sampling design. Sampling designs that maximize independent information maximize cross‐validation bias but also maximize true R². The bagging tree model is found to generally outperform the other regression models in the study on several criteria. Finally, we discuss and recommend use of LiDAR in conjunction with regression modelling to advance understanding of snow depth spatial distribution at spatial scales of thousands of square kilometres. Copyright © 2012 John Wiley & Sons, Ltd.     

A rain on snow climatology and temporal analysis for the eastern United States

ABSTRACT

Rain-on-snow (ROS) has the potential to produce devastating floods by enhancing runoff from snowmelt. Although a common phenomenon across the eastern United States, little research has focused on ROS in this region. This study used a gridded observational snow dataset from 1960–2009 to establish a comprehensive seasonal climatology of ROS for this region. Additionally, different rain and snow thresholds were compared while considering temporal trends in ROS occurrence at four grid cells representing individual locations. Results show most ROS events occur in MAM (March-April-May). ROS events identified with rainfall >1 cm are more frequent near the east coast and events identified with >1 cm snow loss are more common in higher latitudes and/or elevations. Decreasing trends in DJF (December-January-February) ROS events were identified near the coastal areas, with increasing trends in the northern portion of the domain. Significant decreasing trends in MAM ROS are likewise present on a regional scale. Factors playing a role in snowpack depth and rainfall, such as movement of storm tracks in this region, should be considered with future work to discern mechanisms causing the changes in ROS frequency.     

Development of a rain‐on‐snow detection algorithm using passive microwave radiometry

Currently observed climate warming in the Arctic has numerous consequences. Of particular relevance, the precipitation regime is modified where mixed and liquid precipitation can occur during the winter season leading to rain‐on‐snow (ROS) events. This phenomenon is responsible for ice crust formation, which has a significant impact on ecosystems (such as biological, hydrological, ecological and physical processes). The spatially and temporally sporadic nature of ROS events makes the phenomenon difficult to monitor using meteorological observations. This paper focuses on the detection of ROS events using passive microwave (PMW) data from a modified brightness temperature (T_B) gradient approach at 19 and 37 GHz. The approach presented here was developed empirically for observed ROS events with coincident ground‐based PMW measurements in Sherbrooke, Quebec, Canada. It was then tested in Nunavik, Quebec, with the Advanced Microwave Scanning Radiometer for the Earth Observing System (AMSR‐E). We obtained a detection accuracy of 57, 71 and 89% for ROS detection for three AMSR‐E grid cells with a maximum error of 7% when considering all omissions and commissions with regard to the total number of AMSR‐E passes throughout the winter period. Copyright © 2016 John Wiley & Sons, Ltd.     

Analysis of snowpack dynamics during the spring melt season for a sub‐alpine site using point measurements and numerical modeling

Snowpack dynamics through October 2014–June 2017 were described for a forested, sub‐alpine field site in southeastern Wyoming. Point measurements of wetness and density were combined with numerical modeling and continuous time series of snow depth, snow temperature, and snowpack outflow to identify 5 major classes of distinct snowpack conditions. Class (i) is characterized by no snowpack outflow and variable average snowpack temperature and density. Class (ii) is characterized by short durations of liquid water in the upper snowpack, snowpack outflow values of 0.0008–0.005 cm hr^?1, an increase in snowpack temperature, and average snow density between 0.25–0.35 g cm^?3. Class (iii) is characterized by a partially saturated wetness profile, snowpack outflow values of 0.005–0.25 cm hr^?1, snowpack temperature near 0 °C, and average snow density between 0.25–0.40 g cm^?3. Class (iv) is characterized by strong diurnal snowpack outflow pattern with values as high as 0.75 cm hr^?1, stable snowpack temperature near 0 °C, and stable average snow density between 0.35–0.45 g cm^?3. Class (v) occurs intermittently between Classes (ii)–(iv) and displays low snowpack outflow values between 0.0008–0.04 cm hr^?1, a slight decrease in temperature relative to the preceding class, and similar densities to the preceding class. Numerical modeling of snowpack properties with SNOWPACK using both the Storage Threshold scheme and Richards' equation was used to quantify the effect of snowpack capillarity on predictions of snowpack outflow and other snowpack properties. Results indicate that both simulations are able to predict snow depth, snow temperature, and snow density reasonably well with little difference between the 2 water transport schemes. Richards' equation more accurately simulates the timing of snowpack outflow over the Storage Threshold scheme, especially early in the melt season and at diurnal timescales.     

An examination of snow redistribution over smooth land‐fast sea ice

An understanding of temporal evolution of snow on sea ice at different spatial scales is essential for improvement of snow parameterization in sea ice models. One of the problems we face, however, is that long‐term climate data are routinely available for land and not for sea ice. In this paper, we examine the temporal evolution of snow over smooth land‐fast first‐year sea ice using observational and modelled data. Changes in probability density functions indicate that depositional and drifting events control the evolution of snow distribution. Geostatistical analysis suggests that snowdrifts increased over the study period, and the orientation was related to the meteorological conditions. At the microscale, the temporal evolution of the snowdrifts was a product of infilling in the valleys between drifts. Results using two shore‐based climate reporting stations (Paulatuk and Tuktoyuktuk, NWT) suggest that on‐ice air temperature and relative humidity can be estimated using air temperature recorded at either station. Wind speed, direction and precipitation on ice cannot be accurately estimated using meteorological data from either station. The temporal evolution of snow distribution over smooth land‐fast sea ice was modelled using SnowModel and four different forcing regimes. The results from these model runs indicate a lack of agreement between observed distribution and model outputs. The reasons for these results are lack of meteorological measurements prior to the end of January, lack of spatially adequate surface topography and discrepancies between meteorological variables on land and ice. Copyright © 2009 John Wiley & Sons, Ltd.     

Etude isotopique de la pluie et de la neige sur le Mont Liban: premiers résultats / Isotope study of snow and rain on Mount Lebanon: preliminary results

Résumé

Résumé Quelques analyses isotopiques préliminaires ont été réalisées sur les précipitations pluvio-neigeuses, sur un profil de neige et sur deux sources karstiques sur le Mont Liban. Elles confirment la variabilité saisonnière du signal atmosphérique et en particulier que l’excès en deutérium est en relation avec l’origine des masses d’air et avec les recharges de vapeur sur la Méditerranée. Elles montrent également une relative stabilité du signal isotopique du couvert neigeux, peu ou pas influencé par les phénomènes de sublimation, d’évaporation ou de fonte/regel. La participation progressive de la fonte du manteau neigeux à l’alimentation des sources karstiques est qualitativement observée.     

Monitoring over‐winter soil water dynamics in a freezing and snow‐covered environment using a thermally insulated tensiometer

Water potential below a frozen soil layer was continuously monitored over an entire winter period (using thermally insulated tensiometers sheltered in a heated chamber) along with other soil, snow and atmospheric variables. In early winter, the freezing front advanced under a thin snow cover, inducing upward soil water flow in the underlying unfrozen soil. The freezing front started to retreat when the snow cover became thick enough to insulate the soil, resulting in the reversal of the flow direction in the unfrozen zone. These data provide a clear illustration of soil water dynamics, which have rarely been monitored with a tensiometer. Copyright © 2005 John Wiley & Sons, Ltd.     

承德初霜冻预报指标分析及其预报方法

应用常规观测资料与ncep1°×1°再分析资料,对承德市各县区初霜的时空分布特征及南北县区初霜出现时间的稳定性进行了分析,在对117个初霜环流形势分析的基础上,建立了初霜预报的天气概念模型;筛选因子,分县区确立了预报指标;采用概率、指标、数值预报产品相结合的方法及概率区间取值法建立了分县区初霜预报模型;检验表明,该模型对于中重度初霜冻预报准确率可达95%以上,对轻度初霜的预报准确率可达68%以上,无漏报出现。     

阿尔山地区积雪变化特征

选取阿尔山气象站1981—2015年冷季(10月—次年4月)气象资料,利用滑动平均、线性倾向估计和Mann-Kendall等方法,对年最大积雪深度、积雪日数、气温和降水量进行分析。结果表明,阿尔山地区年最大积雪深度主要发生在1月至3月,其中2月份概率最大,达50%;34 a内最大积雪深度呈上升趋势(2.77 cm/10a),年平均增加0.98%,且年最大积雪深度在1998年发生了突变,即在1998年之前增长缓慢,在2000年以后上升趋势显著。积雪日数的统计分析表明,初始积雪日数和有效积雪日数呈现略微减少趋势,而稳定积雪日数有微弱的增加趋势;通常初始积雪日数比有效积雪日数大30天左右。年最大积雪深度与稳定积雪时期的降水量、积雪日数、日照时数有显著的相关性,相关系数分别为0.647、0.515、0.584,但与稳定积雪时期的气温没有明显的相关性。在全球变暖的大环境下,积雪深度随着降水量和日照时数的增加而增加,且积雪深度受降水量的影响大于日照时数的影响。     

武威市初、终霜日气候特征

为了因地制宜,合理调整种植结构和布局,有效利用农业气候资源。利用1961—2014年武威市4个气象站点初、终霜日(最低地温≤0℃)观测资料,采用现代气候诊断分析方法,系统分析了该市年初、终霜日的时空变化特征。结果表明:受海拔高度、地形地势以及植被覆盖情况的影响,在空间分布上,武威市初霜日为山区早于荒漠区早于绿洲平原区,终霜日为山区晚于荒漠区晚于绿洲平原区,各地初、终霜日存在一定的异常性,正常初、终霜日均在60%左右,对农业生产造成危害的偏早和特早初霜日、偏晚和特晚终霜日的概率均在20%左右。在时间变化上,武威市初霜日呈显著推迟趋势,终霜日呈显著提早趋势,霜期呈显著缩短趋势,终霜日提早的幅度比初霜日推迟的幅度更大。初霜日和终霜日的时间序列均分别存在着8~10年和9~11年的准周期变化。初霜日在1998年发生了气候突变,终霜日在1996年发生了气候突变。