Evapotranspiration models for mountain valleys

Evapotranspiration models are of special interest for mountains in semi-arid or arid environments. Water management that relies on the role of water towers (Price & Barry 1997) requires improved hydrological models. Amongst important hydrometeorological parameters such as precipitation and ET, runoff is the only reliably measured regional hydrological component under mountain conditions (Whiting 2003). Of the remaining components, regional precipitation is even more difficult to determine (Sevruk & Martinec 1985) than ET. Therefore, it should be more accurate to solve the regional water balance from the sum of discharge and evaporation for dry pentades or decades as well as on an annual timescale (Molnar et al. 1990, de Jong 2002, Weingartner etal. 2003). However, it is not sufficient to derive evaporation from functions without any further validation. Standard procedures rely on meteorological data and parameters typically available only for the valley bottom. When regionalising these data, parameters are introduced according to assumptions supported only by linear altitudinal regressions. Apparently, all that is required is a suitable DEM and GIS (Geographical Information System) of the vegetation or land cover but when looking at the results in detail, the local water balance requires a lot of improvement and validation.

In contrast to precipitation-runoff models, which are delineated according to the extent of valleys and sub-basins, evaporation models should comprise units with more or less uniform evaporation and can therefore extend across several discharge subunits. These units can be defined from areas with homogeneous vegetation cover and surface properties, including soils, aspect and topography (Hipps & Kustas 2001), as well as geomorphology in areas with, for example, glaciers or scree slopes. One problem concerning regional ET models is that only the horizontal projection and not the actual size of slopes is treated. For example, a steep slope with approximately 75° average inclination and 200 m length is represented by only a quarter of its topographic length in a topographic map of 1:25,000. Appropriate correction functions are therefore desirable. Another handicap for evaporation models is that direct ET measurements with lysimeters in mountains cover a density ofhardly 1/10,000 km2. For validation purposes, there is a strong demand for more observations and measurements on ET and evaporation. As shown by de Jong (2002) and in Chapter 12 of this book, electronic measuring systems are relatively easy to construct and affordable.

In mountain catchments, the extent of local water bodies or areas of potential evaporation can be determined from topographical maps, air photos or satellite images. However, the quantification of the extent of lakes, ponds, mires, creeks and river surfaces is particularly problematic during phases of intense snowmelt or extreme precipitation. The parameterisation of dynamic zones with saturated soil and wet locations on slopes remains difficult. According to Dunne etal. (1975), the resulting saturated zones form narrow bands in steep neighbouring slopes, but they are wide in gently inclined neighbouring slopes. Under these circumstances, subsurface transmissivity and slope length can help in defining the potential zone of saturated soil. This local situation affects vegetation and the amount of organic content in the soils of the related zones.

With these large local variations, no linear decline in ET can be expected between alpine valley bottoms and peaks (de Jong 2002). The anticipated decline of ET with altitude in parallel with a decline in organic matter is typical only for the poor pasture zone lying above the steep valley troughs (usually above 2200 m). Investigations show that the valley bottom has a strong tendency towards cooling, reducing ET even during the summer. A further reduction in ET is caused by the spatial transport and distribution of humidity and condensation. Evaporation and transpiration on the lower trough slopes is more intensive than on the valley bottom or the higher slopes even where alpenrose (Rhododendrum ferrugineum) has replaced the former forest cover (de Jong 2002, Chapter 12 of this book). The variability of radiation with aspect induces variations in the timing of maximum transpiration rates, but the difference between the sum of daily ET on east and west sloping sites remains small.

The local topographic position of a site has a more important influence on ET. The more exposed it is to wind, the higher the ET. Special geomorphic zones such as lateral moraines encourage high ET in relation to the surrounding slopes. On the other hand, units of positive relief do not receive lateral water and are therefore the first to suffer desiccation with reduced summer rainfall. Such sites cool intensively during the winter in the absence of thick snow covers and are more prone to long-lasting soil frost. In the alpine meadow and shrub zones above the tree line (at approximately 2300 m in the Dischma), geomorphological zones can differ substantially. They can range from wet zones typical for the centres of corries or wet niches behind lateral moraines on trough shoulders to relatively dry rock faces and scree slopes (Figure 17.2). Indicators for exceptionally humid zones include mossy vegetation in moors and depressions near local moraines and lichens on boulders or on older scree particles. During snowfall and drift or during snowmelt, micro relief amplifies differences in subsurface humidity and temperature. In general, geomorphological zones that are homogeneous over large areas (see upper Dischma Figure 17.2) are easy to delineate from remotely sensed images (Figure 17.3a) or from good topographical maps.

One of the greatest restrictions for quantifying ET is the unsolved problem of defining local roughness of vegetation, rock or sedimentary surfaces. Even micro surfaces of different rocks vary immensely. They are very large for sandstones or schists but rather small for many limestones. No data is available on the potential of different rock surfaces to temporarily store surface water films. Nevertheless, climbers know that large variabilities exist between different slopes. Vegetation and its potential for interception vary significantly over time. Interactions between snow and vegetation during snowmelt typically induce high evaporation. With their unique geomorphic setting, ''Schneetaelchen'' or ''snow valleys'' (Franz 1979) are of special interest since they allow snow to persist during the short summers and most of the meltwater to evaporate at the outlet. What is typical locally is also true for whole valley slopes: periods and regions with snowmelt correspond with the largest annual evaporation rates. In contrast, rock faces and large scree surfaces evaporate most intercepted rainwater and condensated water from their wetted surface. It is therefore reasonable to assume that about 2 mm of intercepted rain or condensation will evaporate per day and that only surplus water will infiltrate into talus scree.

Evapotranspiration models should be differentiated and scaled according to these observations. Surface descriptions should not be limited to vegetation cover (e.g. different grass types, shrubs, forest or bog) or selected soil properties alone. Under alpine conditions, the properties of these surfaces must be differentiated according to geomorphological zones with dominant

Figure 17.2 (Plate 13) Detailed geomorphological map of the Dischma divided into 14 different zones. The valley is dominated by moraines, scree slopes, rock faces, alluvial fans, glacier and snow fields

Figure 17.3 (Plate 6) (a) Remotely sensed CASI (Canadian Airborne System) image of the Duerrboden mass movement in Upper Dischma with a resolution of 5 m in the infrared canal and (b) enlargement of geomorphological map of corresponding area

Zone of detachment

Zone of accumul.

Rock face

Scree slope

Alpine pasture

Figure 17.3 (Plate 6) (a) Remotely sensed CASI (Canadian Airborne System) image of the Duerrboden mass movement in Upper Dischma with a resolution of 5 m in the infrared canal and (b) enlargement of geomorphological map of corresponding area units such as rock surfaces, scree slopes, trough slopes, ground moraines and terminal moraines. Spring horizons resulting from the lateral movement of interflow should be considered on the lower valley slopes or in areas where the carrying capacity of water is reduced owing to changes in the transmissivity of scree material. For similar reasons, slopes prone to mass movements are hydrologically complex and should be carefully integrated into models.

For validating the spatial and temporal distribution of ET, meteorological variables should be obtained at the regional scale by remote sensing or from infrared images. This is especially true for temperature, which should be scanned regularly to complement multiple climate station profiles and improve our knowledge on the budget of sensible and latent heat.

0 0

Post a comment