Evaporation in the Atacama Desert: An empirical study of spatio-temporal variations and their causes

The Atacama Desert is hyper-arid, and areas where adequate moisture exists for evaporation are spatially highly restricted. Nevertheless, water resources exist and their evaluation requires knowledge of this elusive but important component of the hydrological cycle. Evaporation may occur in four typical areas: rivers and associated riparian zones, localized springs, large playas and extensive areas of bare soil after infrequent precipitation events. Transpiration is locally possible where moisture is sufficiently close to the surface to allow phreatophytes or scarce grass cover to grow, but virtually no information is available for quantification. Pan evaporation data from 11 stations for the period 1977–1991 is analyzed and complemented by analysis of an evaporation study conducted in the Salar de Atacama during 1987/1988. The results show that pan evaporation, and hence maximum potential evaporation may be considered largely a function of maximum temperature and elevation as well as density of the evaporating fluid. Actual evaporation is limited by available moisture and diminishes rapidly as the level of soil moisture saturation drops below the soil surface, extinguishing at ca. 2 m depth. Evaporation is greatest during the summer, but at higher elevations convective cloudiness develops during January and February reducing evaporating rates at a time when significant precipitation may occur. Inter-annual variations in pan evaporation are considerable and weakly correlated with ENSO, but variations in actual evaporation are damped by comparison. Regression equations are developed which have widespread applicability and may be used to estimate evaporation in areas where no site-specific data exists.     

Infiltration-recharge through wadi beds in arid regions

Abstract

Groundwater recharge in arid regions is intermittent and usually occurs as a result of flood flow transmission losses in dry wadi channels. Hydrograph characteristics play a dominant role in determining the amount of channel abstraction in relation to the width of the wetted perimeter and the time of inundation, and the subsequent groundwater recharge. Large variations in the magnitude of channel losses result mainly from the diversity in inflow volumes. The magnitude of groundwater recharge in relation to bed transmission losses is dependent on flood volume and duration, soil moisture content and physical soil profile characteristics. Runoff volume and duration are the dominant factors influencing the cumulative infiltrated volume and recharge to shallow water tables. Taking into consideration the influence of various hydrological and channel characteristics, several regression equations are suggested to estimate the transmission losses from a wadi bed and the groundwater recharge.     

The influence of antecedent rainfall on Sahelian surface evaporation

This study quantifies the influence of rainfall on surface evaporation in the Sahel. A numerical model of the surface is used to extend the observations taken during the HAPEX–Sahel project, and is forced by 2 years of rain‐gauge data. The model is applied to the Southern Super Site (SSS), which covers an area of less than 100 km². The effects of rainfall variability (spatial and temporal) on soil moisture, vegetation growth and evaporation are explored. Contrasting rainfall conditions between the two years produce observed differences in the timing of the seasonal growth cycle. This correlates well with modelled root‐zone moisture deficits, and exerts a modest influence on transpiration rates. The evolution of surface evaporation is dominated, however, by the bare soil contribution in the day or two after a storm. This component also exerts a strong influence on the spatial variability of fluxes across the SSS, particularly when rain falls only in part of the area. In these cases, differences in evaporation between recently wetted and dry areas can reach 3\5 mm day⁻¹. Observations indicate that during a period of persistent rainfall gradients across the SSS, the lower atmosphere maintained a ‘memory’ of past rainfall patterns through humidity contrasts. These were the result of gradients in surface soil moisture, and therefore evaporation. The model results therefore support the possibility of a positive surface feedback mechanism affecting rainfall patterns in the region. Copyright © 2000 John Wiley & Sons, Ltd.     

Evapotranspiration from the margin and moist playa of a closed desert valley

Two long term microclimate measurement stations with Bowen ratio capability have been used to study water cycling in a closed desert basin. Microclimate variables including the temperature and vapor pressure gradients were monitored continuously and were used to estimate the Bowen ratio, sensible and latent heat fluxes during 1986 and 1987. Despite having a water table that varied between the surface and 30 cm below the surface, the playa had little evaporation except after rainfall events. The very high osmotic pressure of the soil and salt crust caused most of the absorbed radiation to be partitioned to sensible heat. In contrast, along the margin the thin grass and brush cover transpired water freely, with the latent heat flux exceeding 60% of available energy for much of the season. The higher air temperatures above the playa raised potential evapotranspiration (ET) significantly higher than along the margin throughout the summer. The annual average actual ET of the playa was only 36% of the margin. During the drier summer period (May–October), this ratio decreased to < 28%. Immediately after a rainfall event, evaporation rates of the two sites were similar, but the playa rate was quickly reduced as the osmotic potential increased. During this study, the playa lost < 229 mm of subsurface water to evaporation annually, while > 638 mm were lost from the margin groundwater supply.

The 24-h solar and net radiation correlations were 0.80 and 0.94 for the playa and margin, respectively. The lower correlation for the playa resulted from the wide variation of albedo with surface moisture changes. The annual average albedo values for the playa and margin were 0.64 and 0.46, respectively.     

An objective assessment of soil-moisture deficit models

The performance of different soil-moisture deficit models was assessed by comparing model predictions with over 3000 neutron-probe observations of soil-moisture content at six grassland experimental sites operated by the Institute of Hydrology. The models were formulated using combinations of different equations for estimating potential evaporation and different regulating functions relating actual to potential evaporation via the moisture status of the soil. The inclusion of sophisticated evaporation equations (Priestley-Taylor, Penman, Thom-Oliver) gave no improvement in SMD prediction over a proposed simple evaporation formula requiring no direct meteorological measurements other than rainfall. The success of this formula demonstrates the conservative nature of annual potential evaporation within the U.K. both spatially and temporally and also suggests a possible natural feedback mechanism between atmospheric demand and grass transpiration.     

Estimates of spatial variation in evaporation using satellite‐derived surface temperature and a water balance model

Evaporation dominates the water balance in arid and semi‐arid areas. The estimation of evaporation by land‐cover type is important for proper management of scarce water resources. Here, we present a method to assess spatial and temporal patterns of actual evaporation by relating water balance evaporation estimates to satellite‐derived radiometric surface temperature. The method is applied to a heterogeneous landscape in the Krishna River basin in south India using 10‐day composites of NOAA advanced very high‐resolution radiometer satellite imagery. The surface temperature predicts the difference between reference evaporation and modelled actual evaporation well in the four catchments (r² = 0·85 to r² = 0·88). Spatial and temporal variations in evaporation are linked to vegetation type and irrigation. During the monsoon season (June–September), evaporation occurs quite uniformly over the case‐study area (1·7–2·1 mm day^?1), since precipitation is in excess of soil moisture holding capacity, but it is higher in irrigated areas (2·2–2·7 mm day^?1). In the post‐monsoon season (December–March) evaporation is highest in irrigated areas (2·4 mm day^?1). A seemingly reasonable estimate of temporal and spatial patterns of evaporation can be made without the use of more complex and data‐intensive methods; the method also constrains satellite estimates of evaporation by the annual water balance, thereby assuring accuracy at the seasonal and annual time‐scales. Copyright © 2007 John Wiley & Sons, Ltd.     

Modification of the Penman method for computing bare soil evaporation

The Penman equation, which calculates potential evaporation, was modified by Staple (1974, Soil Science Society of America Proceedings 38 : 837) to include in it the relative vapour pressure h_s of an unsaturated soil to predict actual evaporation from a soil surface. This improved the prediction when the difference between the temperature of the soil surface and ambient air is relatively small. The objectives of this study were (i) to revise it further using the actual temperature of the soil surface and air to provide the upper boundary condition in computing evaporative flux from the soil surface and (ii) to determine the range of water content for which the modified form of the Penman equation is applicable. The method adopted was tested by a series of outdoor experiments with a clay soil. The method of Staple (1974) overestimated the rate of evaporation above the water content 0·14 m³ m^?3 (up to 30% deviation), whereas the new method agreed well with the measured rates (maximum 7% deviation). Below 0·14 m³ m^?3 water content, both methods underestimated, but the Staple (1974) method deviated more from the measured values: the deviations were above 70% and around 30% for the Staple (1974) and the new methods respectively. Although the new method provided accurate solutions for a wider range of water content from saturation to the lower limit of the liquid phase of a particular soil, the modification did not respond to the vapour phase of the soil moisture. Therefore, in the dry range (i.e. in the vapour phase in which the flow was entirely as vapour), either resistance models or a Fickian equation should be used. Although the effect of salinity on the measured rates was significant, the model erroneously calculated the same rates for both saline and non‐saline conditions. The effect of soil texture can easily be accounted by defining appropriate matric potential water content ψ_m(θ) and soil relative humidity water content h_s(θ) relationships. Copyright © 2007 John Wiley & Sons, Ltd.     

Observation and estimation of evaporation from the ground surface of the cryosphere in eastern Asia

The characteristics of evaporation from the ground surface of Asian cryosphere sites are presented, as estimated by the lysimeter method, a profile method, and a heat budget method. The observation sites were located on the eastern Tibetan Plateau, in the Qilian and Tianshan Mountains of China, and in eastern Siberia. The lysimeter method has been demonstrated to be a reliable observation technique for estimating daily evaporation from the land surface, given suitable experiment design and operation. Daily mean evaporation varied within the range of 0·3 to 3·5 mm on the permafrost surface, and regional differences in evaporation were strongly related to surface soil moisture. Locally, topography, by way of its influence on surface soil moisture, was found to control evaporation systematically. Seasonality of ground evaporation in permafrost regions is dominated by thaw–freeze cycles at the surface; evaporation from the melting permafrost surface is up to four to seven times greater than that from frozen ground. In forested terrain, the interception of precipitation can reduce daily evaporation by 60 to 70%. Sublimation from the snow surface was observed at some sites in the range of 0·2 to 1·0 mm daily; atmospheric conditions, such as wind speed and saturation deficit, were dominant factors in determining snow sublimation. Copyright © 2003 John Wiley & Sons, Ltd.     

Evaporation Process in Soil Surface Containing Calcic Nodules on the Northern Loess Plateau of China by Simulated Experiments

Soil containing calcic nodules is widely present on the northern Loess Plateau of China owing to soil genesis under local climate conditions. In most studies, little attention is payed to the effect of calcic nodules on soil evaporation and ecoenvironment, resulting in inaccurate evaporation estimation in this kind of soil and further improper field water management measures and irrigation effects. In this paper, soil column experiments were conducted in order to investigate evaporation process in soil containing calcic nodules and the effect of calcic nodules on soil evaporation was determined. The results indicated that evaporation reduction was positively related to calcic nodule content (CNC = mass of calcic nodules/total mass), and could be estimated by the experiential equation: E_soil = E₀ (1 – 0.4 CNC) (E_soil = actual evaporation, E₀ = theory evaporation in soil without calcic nodules). When CNC was below 0.2, the impact could be neglected. While, as CNC exceeded 0.2, the impact needed to be considered during soil evaporation estimation. As CNC reached 0.5, soil evaporation could be reduced by 7.5 mm, accounting for around 10% of the total soil water. Water balance calculation in soil columns showed that water absorbed by calcic nodules was partially available to evaporation. Water available to evaporation was positively related to CNC, and this water could not exceed 63% of the water absorbed by calcic nodules. Generally, evaporation behavior was dominated by calcic nodule quantity and its water absorption. These results provide new ideas for irrigation measures in arid areas of the globe.     

Catchment evaporation and potential evaporation further development of a climatologic relationship

When evaporation from the surfaces of a region is limited by the availability of moisture the resultant changes in the temperature and humidity of the overpassing air are reflected in the potential evaporation. Such interactions have provided the basis for the progressive development of a model in which the regional evaporation, a product of climatic, soil moisture and vegetative processes, is estimated from its effects on potential evaporation or on representative climatological observations. Further improvements are presented and the results are tested with previously published hydrologic and climatologic data from twenty river catchments in Canada and Ireland. These catchments have a range in annual runoff of approximately zero to one meter and a range of snow cover duration from almost zero to six months. Comparison of model predictions with precipitation less runoff estimates for five year periods provides reasonable evidence for the validity of the model with a maximum deviation of 44 mm yr⁻¹, or 9%, and a standard error of estimate of 26 mm yr⁻¹. Seasonal comparisons permit insights into the variability of catchment storage and the effects of climate and topograpy on regional evaporation.     

Simulation of willow short‐rotation forest evaporation using a modified Shuttleworth–Wallace approach

Evaporation from a willow short‐rotation forest was analysed using a modified version of the Shuttleworth–Wallace model. The main modification consisted of a two‐layer soil module, which enabled soil surface resistance to be calculated as a function of the wetness of the top soil. Introduction of the threshold value of the leaf area index when scaling up from the leaf to the canopy resistance resulted in improvement to the simulated evaporation. The analysis was concentrated mainly on the 1988 season (May–October) when total evaporation was measured by the energy balance/Bowen ratio method throughout the growing season, covering all stages of canopy development. At the beginning of the 1994 season, soil evaporation were also measured with a ventilated chamber system. The general seasonal dynamics of the evaporation were fairly well simulated with the model. The largest deviation between measured and simulated evaporation occurred in June, when the model underestimated evaporation by about 1 mm day^?1. The model underestimated also in May but not as much as in June. In September and October the performance of the model was very good. For 130 days of the period May–October the cumulated measured evaporation was 364 mm and the simulated evaporation for the same days was 362 mm. It should be pointed out that this result was obtained without calibrating the model against the measured evaporation. The total simulated evaporation for the season was 450 mm with transpiration constituting 298 mm (66%), soil evaporation 102 mm (23%) and interception evaporation 50 mm (11%). The sensitivity analysis showed, in general, that simulated evaporation was most sensitive to changes in resistances when the leaf area index was smallest, i.e. under non‐closed canopy conditions. Changes in stomatal resistance, which is one of the most sensitive parameters, with associated changes in canopy transpiration, resulted in a negative feedback effect on soil evaporation. This reduced the total evaporation's sensitivity to stomatal resistance. This type of interaction between canopy and soil or undergrowth fluxes has been observed in other studies as well. Copyright © 2001 John Wiley & Sons, Ltd.     

The effects of possible future climate change on evaporation losses from four contrasting UK water catchment areas

Evaporation losses from four water catchment areas under different land uses and climatic conditions were calculated using formulations developed from small plot studies. These formulations, dependent on rainfall inputs, potential evaporation and air temperature, were extrapolated to the catchment scale using land classifications based on analysing remotely sensed imagery. The approach adopted was verified by comparing the estimated annual evaporation losses with catchment water use, given by the difference between rainfall inputs and stream flow outputs, allowing for changes in soil moisture. This procedure was repeated using modified values of rainfall, potential evaporation and air temperature, as given by a climate change scenario. The computed evaporation losses were used in annual water balances to calculate stream flow losses under the climate change scenario. It was found that, in general, stream flow from areas receiving high rainfall would increase as a result of climate change. For low rainfall areas, a decrease in stream flow was predicted. The largest actual changes in stream flow were predicted to occur during the winter months, although the largest percentage changes will occur during the summer months. The implications of these changes on potable water supply are discussed. © 1998 John Wiley & Sons, Ltd.     

Modelling non-stationary water ages in a tropical rainforest: A preliminary spatially distributed assessment

Pristine tropical forests play a critical role in regional and global climate systems. For a better understanding of the eco-hydrology of tropical “evergreen” vegetation, it is essential to know the partitioning of water into transpiration and evaporation, runoff and associated water ages. For this purpose, we evaluated how topography and vegetation influence water flux and age dynamics at high temporal (hourly) and spatial (10 m) resolution using the Spatially Distributed Tracer-Aided Rainfall-Runoff model for the tropics (STARRtropics). The model was applied in a tropical rainforest catchment (3.2 km²) where data were collected biweekly to monthly and during intensive monitoring campaigns from January 2013 to July 2018. The STARRtropics model was further developed, incorporating an isotope mass balance for evapotranspiration partitioning into transpiration and evaporation. Results exhibited a rapid streamflow response to rainfall inputs (water and isotopes) with limited mixing and a largely time-invariant baseflow isotope composition. Simulated soil water storage showed a transient response to rainfall inputs with a seasonal component directly resembling the streamflow dynamics which was independently evaluated using soil water content measurements. High transpiration fluxes (max 7 mm/day) were linked to lower slope gradients, deeper soils and greater leaf area index. Overall water partitioning resulted in 65% of the actual evapotranspiration being driven by vegetation with high transpiration rates over the drier months compared to the wet season. Time scales of water age were highly variable, ranging from hours to a few years. Stream water ages were conceptualized as a mixture of younger soil water and slightly older, deeper soil water and shallow groundwater with a maximum age of roughly 2 years during drought conditions (722 days). The simulated soil water ages ranged from hours to 162 days and for shallow groundwater up to 1,200 days. Despite the model assumptions, experimental challenges and data limitation, this preliminary spatially distributed model study enhances knowledge about the water ages and overall young water dominance in a tropical rainforest with little influence of deeper and older groundwater.     

Seasonal ground ice impacts on spring ecohydrological conditions in a western boreal plains peatland

Peatlands in the Western Boreal Plains act as important water sources in the landscape. Their persistence, despite potential evapotranspiration (PET) often exceeding annual precipitation, is attributed to various water storage mechanisms. One storage element that has been understudied is seasonal ground ice (SGI). This study characterized spring SGI conditions and explored its impacts on available energy, actual evapotranspiration, water table, and near surface soil moisture in a western boreal plains peatland. The majority of SGI melt took place over May 2017. Microtopography had limited impact on melt rates due to wet conditions. SGI melt released 139mm in ice water equivalent (IWE) within the top 30cm of the peat, and weak significant relationships with water table and surface moisture suggest that SGI could be important for maintaining vegetation transpiration during dry springs. Melting SGI decreased available energy causing small reductions in PET (<10mm over the melt period) and appeared to reduce actual evapotranspiration variability but not mean rates, likely due to slow melt rates. This suggests that melting SGI supplies water, allowing evapotranspiration to occur at near potential rates, but reduces the overall rate at which evapotranspiration could occur (PET). The role of SGI may help peatlands in headwater catchments act as a conveyor of water to downstream landscapes during the spring while acting as a supply of water for the peatland. Future work should investigate SGI influences on evapotranspiration under differing peatland types, wet and dry spring conditions, and if the spatial variability of SGI melt leads to spatial variability in evapotranspiration.     

Estimating soil water fluxes from soil water records obtained using dielectric sensors

Accurate estimation of the soil water balance (SWB) is important for a number of applications (e.g. environmental, meteorological, agronomical and hydrological). The objective of this study was to develop and test techniques for the estimation of soil water fluxes and SWB components (particularly infiltration, evaporation and drainage below the root zone) from soil water records. The work presented here is based on profile soil moisture data measured using dielectric methods, at 30‐min resolution, at an experimental site with different vegetation covers (barley, sunflower and bare soil). Estimates of infiltration were derived by assuming that observed gains in the soil profile water content during rainfall were due to infiltration. Inaccuracies related to diurnal fluctuations present in the dielectric‐based soil water records are resolved by filtering the data with adequate threshold values. Inconsistencies caused by the redistribution of water after rain events were corrected by allowing for a redistribution period before computing water gains. Estimates of evaporation and drainage were derived from water losses above and below the deepest zero flux plane (ZFP), respectively. The evaporation estimates for the sunflower field were compared to evaporation data obtained with an eddy covariance (EC) system located elsewhere in the field. The EC estimate of total evaporation for the growing season was about 25% larger than that derived from the soil water records. This was consistent with differences in crop growth (based on direct measurements of biomass, and field mapping of vegetation using laser altimetry) between the EC footprint and the area of the field used for soil moisture monitoring. Copyright © 2007 John Wiley & Sons, Ltd.     

Regression method for estimating rainfall recharge at unconfined sandy aquifers with an equatorial climate

A reliable estimate of rainfall recharge is essential for groundwater system managements. This study develops a method based on regression equations for estimating rainfall recharge at unconfined sandy aquifers with an equatorial climate. The developed method (GR-I method) is generally efficient for estimating long-term regional recharge, as the computational procedures could be formulated and executed easily using Microsoft's Excel spreadsheet. More importantly, its application could be extended to sand textures different from the sand texture used in developing the regression equations. To evaluate its reliability, the method was applied to estimate monthly gross recharge percentages at the Changi reclaimed land. When ignoring the effect of rainfall clusters, the GR-I method was found to underestimate the monthly gross recharge percentages for those months with high monthly rainfall depths. By integrating the effect of rainfall clusters, the GR-I method yields reliable estimates of monthly gross recharge percentages. By including daily potential evaporation as an additional input variable, the Extended GR-I method was found to be not superior to the GR-I method, implying that soil moisture availability is the major governing factor for actual soil evaporation in the highly porous sand medium, instead of atmospheric demand represented by the potential evaporation rate. Using the GR-I method, the mean annual net recharge percentage of the study site was found to fall between 56·9 and 69·9%, which corresponds to a net recharge depth of 1073·8–1745·8 mm. Although the developed method provides a good alternative to other widely used methods, its recharge estimates still needs to be collaborated with estimates from other methods, as multiple techniques are highly recommended in any groundwater recharge estimations. Copyright © 2007 John Wiley & Sons, Ltd.     

On the role of soil moisture in daytime evolution of temperatures

This paper explores the relationship between temperature, evaporation and soil moisture using a planetary boundary layer (PBL) model. It focuses on illustrating and quantifying the effect of soil moisture on the evolution of daytime temperatures. A simple convective PBL model coupled with the Penman–Monteith (PM) equation is used to estimate evapotranspiration. Following calibration and sensitivity analysis, the model was used to simulate the relative impact of dry and wet soil moisture conditions on daytime temperatures by changing the surface resistance parameter in the PM equation. It was found that the maximum temperature that can be reached during a day is constrained by the amount of soil moisture and the available net radiation, confirming previously published results. Higher temperatures can be reached with greater net radiation and dry soil moisture conditions. Copyright © 2012 John Wiley & Sons, Ltd.     

Effects of soil moisture aggregation on surface evaporative fluxes

The effects of small-scale heterogeneity in land surface characteristics on the large-scale fluxes of water and energy in the land-atmosphere system has become a central focus of many climatology research experiments. The acquisition of high resolution land surface data through remote sensing and intensive land-climatology field experiments (like HAPEX, FIFE, and BOREAS) has provided data to investigate the interactions between microscale land-atmosphere interactions and macroscale models. To determine the effect of small scale heterogeneities, the spatially averaged evaporative fraction is analytically derived for spatially variable soil moisture and soil-atmospheric controls on evaporation at low soil moisture. This average evaporative fraction is compared with the evaporative fraction determined using the spatially averaged soil moisture, as if from a lumped, or aggregated, land surface model. Results show that the lumped-model based evaporation will over estimate evaporation during periods of low atmospheric demands (early morning/late afternoon, Winter periods, etc.) and under estimate evaporation during periods of high demand (midday Summer periods.) The accuracy of using ‘effective’ parameters in lumped macroscale models depends on the variability of soil moisture and the sensitivity of the soil-vegetation system to low soil moisture.     

Utilization of satellite data in land surface hydrology: sensitivity and assimilation

This paper investigates the sensitivity of potential evapotranspiration to input meteorological variables, i.e. surface air temperature and surface vapor pressure. The sensitivity studies have been carried out for a wide range of land surface variables such as wind speed, leaf area index and surface temperatures. Errors in the surface air temperature and surface vapor pressure result in errors of different signs in the computed potential evapotranspiration. This result has implications for use of estimated values from satellite data or analysis of surface air temperature and surface vapor pressure in large‐scale hydrological modeling. The comparison of cumulative potential evapotranspiration estimates using ground observations and satellite observations over Manhattan, Kansas for a period of several months shows a variable difference between the two estimates. The use of satellite estimates of surface skin temperature in hydrological modeling to update the soil moisture using a physical adjustment concept is studied in detail, including the extent of changes in soil moisture resulting from the assimilation of surface skin temperature. The soil moisture of the 1 cm surface layer was adjusted by 0·9 mm over a 10‐day period as a result of a 3 K difference between the predicted and the observed surface temperature. This is a considerable amount given the fact that the top layer can hold only 5 mm of moisture. Copyright © 2001 John Wiley & Sons, Ltd.     

Validity of the Bouchet’s complementary relationship at 102 observatories across China

Bouchet in 1963 hypothesized that for large homogeneous land surface with minimum advection of heat and moisture, there exists a 1:1 complementary relationship of potential and actual evaporation coupled through land-atmosphere feedbacks. The complementary relationship has been widely used to estimate regional actual evaporation and explain the pan evaporation paradox. We examine the standardized potential evaporation (potential evaporation divided by wet environment evaporation) at 102 observatories at different elevations across China. Generally, the relationship is appropriate at the low elevations (<1000 m). With the increase of elevation, vapor transfer power becomes much less than radiation energy budget because of lower vapor pressure deficit and stronger global solar radiation. As a result, at the high elevations (over 1000 m), the excess energy resulted by limited moisture availability is not enough to be converted into drying power of the air. This result suggests that the complementary relationship is asymmetric at the high elevations. Supported by the Presidential Special Award Foundation, the Chinese Academy of Sciences (Grant No. O7R70020SD) and the National Key Technology R & D Program (Grant No. 2006BAC08B0408)