Improved snow cover model in terrestrial ecosystem models over the Qinghai–Tibetan Plateau

https://doi.org/10.1016/j.agrformet.2015.12.004Get rights and content

Highlights

  • Current models overestimates snow cover days and depth in Qinghai–Tibetan Plateau.

  • New model integrating wind speed improves the performance.

  • Parameter inversion is improved by separating snow accumulation and melting.

Abstract

Snow cover is an important land surface variable, especially in the Qinghai–Tibetan Plateau, and profoundly impacts plant phenology, growth, soil respiration and net ecosystem production. However, few studies have been conducted to examine the reliability of current snow cover models in the Qinghai–Tibetan Plateau. This study examined three snow cover models (SnowMAUS, revised SnowMAUS, and SnowFrostIce) across 23 meteorological stations covering a variety of altitudes. All three models substantially overestimated snow cover depth and numbers of days with snow cover because they ignored the impact of wind speed. Based on SnowMAUS, this study has incorporated equations dealing with wind speed impacts to develop a new model (SnowWind). Moreover, the parameter inversion method was improved by separating snow cover calculations into two processes (snow accumulation and melting) and estimating the parameters of the two processes separately. The results showed that SnowWind can estimate snow cover dynamics with reasonable precision of snow cover presence or absence and, to a large extent, snow cover depth. The better inversion method also improved the number of constrained parameters and model performance. Regional estimates of snow cover are comparable with those derived from satellite-based snow cover products with regard to snow cover days. This result shows the SnowWind model is an alternative approach that makes it possible to simulate sow cover and depth over large areas.

Introduction

Snow cover has been found to be a critical component of terrestrial ecosystem and has a strong influence on carbon and water fluxes between atmosphere and land surface by altering soil temperature, moisture, roughness, and albedo (Stewart et al., 2004). Snow cover regulates the ecosystem water cycle by resisting moisture recharge and recharging soil moisture during thaws (Shanley and Chalmers, 1999). The insulating properties of snow cover substantially impact the processes of soil thawing and freezing (Mellander et al., 2004). Moreover, some studies showed that interannual variations in vegetation growth were primarily governed by year-to-year variations in the length of the snow-free period (Ryo, 2014, Choler, 2015). Numerous studies have highlighted that snow cover dynamics need to be simulated and incorporated into ecosystem models to provide an adequate assessment of the role of snow cover in terrestrial ecosystems (Douville et al., 1995, Zhuang et al., 2001, Walter et al., 2005, Trnka et al., 2010).

Many snow cover models have been developed and integrated into various models (Schlosser et al., 2000, Franz et al., 2008, Trnka et al., 2010, Jégo et al., 2014). Based on the observations from an alpine site, a recent comparison of 1701 snow models showed no best model if consideration of errors in simulations of snow mass, snow depth, albedo and surface temperature (Essery et al., 2013). More importantly, most of these modeling studies were conducted in high-latitude regions (Henderson-Sellers et al., 1995, Bowling et al., 2003, Nijssen et al., 2003), with few model studies in the Qinghai–Tibetan Plateau region. In such a low-latitude alpine region, the land surface and climate properties are different from those at high latitudes (Qin et al., 2006). The Qinghai–Tibetan Plateau is characterized by a semiarid climate with mean annual precipitation below 450 mm, as well as relatively little snowfall (Wang et al., 2009). The dominant plants in the Qinghai–Tibetan Plateau are very low, with a weak capability to block wind from the land surface (Li et al., 2006). Therefore, snow cover models need to be further examined and developed for use over the plateau.

Despite the fact that snow cover studies over the Qinghai–Tibetan Plateau are few, snow cover plays an important role in regulating major ecosystem structure and functions. Previous study showed snow melting date and vegetation green-up date were significantly correlated over large areas of the Qinghai–Tibetan Plateau (Shen, 2014). Seasonal snow cover durations substantially impacted vegetation growth over the Qinghai–Tibetan Plateau, and especially March–May snow cover dynamics played a significantly role in vegetation growth (Wang et al., 2014). Moreover, snow depth affected species composition of the Qinghai–Tibetan Plateau, and the highest species richness and species diversity occurred with an intermediate snow depth, showing a unimodal curve with the increase in snow depth (Chen et al., 2008).

The overarching goal of this study is to examine and improve a snow cover model over the Qinghai–Tibetan Plateau. In particular, the model should adequately incorporate the unique processes over the Qinghai–Tibetan Plateau, and model formulations should rely only on meteorological data that are commonly included in ecosystem models. Specific objectives are to (1) examine the accuracy of current snow cover models in the Qinghai–Tibetan Plateau, (2) improve snow cover models by considering the unique characteristics of the Qinghai–Tibetan Plateau, and (3) compare snow cover predictions with observations from meteorological stations and satellite products.

Section snippets

SnowMAUS model

The SnowMAUS model was originally developed by Running and Coughlan (1988) and modified and examined by Thornton et al. (2000) and Trnka et al. (2010). The SnowMAUS model operates with a daily time step and simulates three major processes of snow cover dynamics: snow accumulation, melting, and sublimation. The snow accumulation (SnowAccu, mm) process separates precipitation (Precip, mm) into snow and rain using a simple linear function of daily minimum air temperature (Tmin, °C):ifTminT

Site comparison of snow cover models

All three previous models (SnowMAUS, revised SnowMAUS, and SnowFrostIce) showed similar performance for simulating snow cover dynamics, with low r and high predictive bias (PE and RMSE). Over all validation sites, on average, the correlation coefficients (r) of the three models were 0.30–0.49 and 0.59–0.71 for snow cover days and snow depth respectively (Fig. 2a and d). All three models overestimated the days and depth of snow cover at almost all sites. On average, the three models

Discussion

Simple snow cover models have been developed by a number of researchers; however, most of these models were used either in mountainous terrain or at high latitudes (Trnka et al., 2010, Jégo et al., 2014). None of them has been examined in the Qinghai–Tibetan Plateau. The results obtained here showed poor performance for three previous models which ignored the impact of wind speed (Fig. 2, Fig. 3). The Qinghai–Tibetan Plateau has unique characteristics of vegetation and climate condition. Except

Summary

This study aimed to examine the reliability of previous snow cover models in the Qinghai–Tibetan Plateau, which is characterized by high wind speeds and flat terrain. The results obtained here showed that three previous snow cover models which were validated and used in other regions overestimated snow cover depth and the number of days with snow because they ignored the impact of wind speed. Based on the SnowMAUS model, a new model (SnowWind) was developed in this research by incorporating

Acknowledgements

This study was supported by Key Project of Chinese Academy of Sciences (CAS) (KJZD-EW-G03-04), One Hundred Person Project of CAS, and National Science Foundation for Excellent Young Scholars of China (41322005). Snow depth dataset derived passive microwave remote-sensing data is provided by Cold and Arid Regions Sciences Data Center at Lanzhou (http://westdc.westgis.ac.cn). We have complied with AGU's Data Policy by providing information on how to obtain the data used to produce the results of

References (43)

  • M.G. Shen

    Increasing altitudinal gradient of spring vegetation phenology during the last decade on the Qinghai–Tibetan Plateau

    Agric. For. Meteorol.

    (2014)
  • P.E. Thornton et al.

    Simultaneous estimation of daily solar radiation and humidity from observed temperature and precipitation: an application over complex terrain in Austria

    Agric. For. Meteorol.

    (2000)
  • M. Trnka

    Simple snow cover model for agrometeorological applications

    Agric. For. Meteorol.

    (2010)
  • G.X. Wang et al.

    The influence of freeze-thaw cycles of active soil layer on surface runoff in a permafrost watershed

    J. Hydrol.

    (2009)
  • W.P. Yuan et al.

    Improving model parameter estimation using coupling relationships between vegetation production and ecosystem respiration

    Ecol. Model.

    (2012)
  • R.G. Allen et al.

    Crop evapotranspiration: guidelines for computing crop water requirements

    (1998)
  • G. Bélanger et al.

    Climate change and winter survival of perennial forage crops in eastern Canada

    Agron. J.

    (2002)
  • T. Che et al.

    Snow depth derived from passive microwave remote-sensing data in China

    Ann. Glaciol.

    (2008)
  • W.N. Chen

    Effect of snow-cover duration on plant species diversity of alpine meadows on the eastern Qinghai–Tibetan Plateau

    J. Mt. Sci.

    (2008)
  • P. Choler

    Growth response of temperate mountain grasslands to inter-annual variations in snow cover duration

    Biogeosciences

    (2015)
  • M. Decker et al.

    Evaluation of the reanalysis products from GSFC, NCEP, and ECMWF using flux Tower observations

    J. Clim.

    (2012)
  • Cited by (24)

    • Scale effect of climate factors on soil organic carbon stock in natural grasslands of northern China

      2023, Ecological Indicators
      Citation Excerpt :

      For example, annual maximum temperature (AMAT) and annual minimum temperature (AMIT) can affect the activity of phosphatase (Jiao et al., 2011), while mean annual humidity (MAH) can affect the activity of catalase (Jiao et al., 2011), and combine with mean annual ground temperature (MAGT) to affect the activity of ectomycorrhizal fungi (Castaño et al., 2017) and surface energy budgets (Shati et al., 2018). Annual sunshine duration (ASD), the number of snow-covered days (NSD) and annual maximum depth of accumulated snow (AMDS) can affect plant photosynthesis and transpiration (Bansouleh et al., 2009), plant phenology (Yuan et al., 2016), and nutrient transformation (Edwards et al., 2007). Therefore, to accurately distinguish the main influencing factors that regulate SOCS in different grassland types, it is essential to select multiple climate factors for analysis and comparison.

    • Spatiotemporal variations of global terrestrial vegetation climate potential productivity under climate change

      2021, Science of the Total Environment
      Citation Excerpt :

      In addition, the Tmx values in the Southern Hemisphere changes greatly, increasing by more than 2.5 °C. Furthermore, the rising global Tmx mainly increases CPP in the high latitude regions, which may be attributed to a more important role of TA in explaining CPP variability (Nemani et al., 2003; Yuan et al., 2016). The increase in Tmx can promote the increase in TA and alleviate the inhibition of low temperatures on the CPP to some extent.

    • Satellite-observed solar-induced chlorophyll fluorescence reveals higher sensitivity of alpine ecosystems to snow cover on the Tibetan Plateau

      2019, Agricultural and Forest Meteorology
      Citation Excerpt :

      And monthly mean snow depth (cm) data from 2007 to 2014 are also collected from these stations. The data have been widely used in many previous studies (Yuan et al., 2016; Xu et al., 2017). The TP in this study is defined as the area with an altitude greater than 3000 m and within the area from 27 to 40°N and 90–105°E covering the eastern part of the TP, where vegetation occurs.

    • Valuating service loss of snow cover in Irtysh River Basin

      2019, Advances in Climate Change Research
      Citation Excerpt :

      The snow depth products have been extensively validated (Dai et al., 2015) and proved to be appropriate to determine peak accumulation and seasonal SWE cycle. Therefore, this product has been widely utilized in the climatic and hydrological research communities (Li et al., 2008; Zheng et al., 2009; Ji and Kang, 2013; Yuan et al., 2016). In economic analysis, the value of natural resources can be categorized into use value and nonuse value.

    View all citing articles on Scopus
    View full text