Climate maps

For the creation of climate maps different state-of-the-art geostatistical interpolation methods were applied with the aid of programming software and geographical information system software. A digital elevation model formed the working basis, from which the spatial resolution of 500×500 m was passed on the resultant maps in the form of digital grid fields. Three main groups of interpolation methods were employed dependent on the climate parameter under consideration. The choice of the method, most appropriate for a certain climate parameter, regards the physical-spatial behaviour of the parameter as well as the statistical functionality of the method.

 

The maps of temperature, snow and fresh snow climate were created applying a recently published method (Frei 2014; Hiebl and Frei 2015), which was developed specifically for the interpolation of temperature in mountainous study regions. It combines a macroclimatic background field, representing large-scale horizontal and basin-scale vertical variations, with a mesoclimatic residual field, representing more local phenomena such as valley-scale cold-pools and warm föhn air layers in Alpine valleys. The background field is constructed from a large-scale vertical profile, varying smoothly in the horizontal and allowing for nonlinear vertical dependence. The vertical profiles are determined for predefined but gradually overlapping subregions of the study domain. The residual field is constructed by weighting the station residuals from the background field. The weighting scheme is not strictly distance-related (Euclidean), but uses a predefined set of generalised distance fields determined from a non-Euclidean distance metric that takes account of the topographic obstruction of air-flows.

 

The maps of precipitation climate were constructed using geographically weighted regressions (Daly et al. 1994; 2008; Frei and Schär 1998). The method strongly involves elevation dependencies and allows for a high degree of spatial variability. However it is sensitive to sparse station coverage in high-elevation areas. An individual regression of the climate parameter against elevation is calculated at each grid point of the target grid involving the closest stations. To control the stations' contribution they are weighted according to their representativeness for the topographical conditions at the grid point. Checking different weights and parameterisations, a specific number of stations is selected to be weighted by their horizontal distance and by their location north or south to the main Alpine crest.

The maps of radiation climate were extracted from an already existing radiation dataset (Olefs and Schöner 2012; Olefs 2013). This dataset involves hourly analyses of near-surface global solar radiation values over the entire study period and accounts for the complex interaction of radiation and topography. Due to the small number of available stations with global radiation data during the 1980s, the latter was calculated from sunshine duration using the Ångstrom formulae and then split into a direct and diffuse part using measurements of global and diffuse radiation as well as relative sunshine duration records. To include the effect of cloudiness difference raster were derived separately for the direct and diffuse component between clear sky values of a solar radiation model and ground measurements.

 

The interpolation performance is evaluated by systematic leave-one-out cross-validation, i.e. the sequential omission of each station observation and its independent prediction by the method. It produces interpolation errors (in terms of mean absolute errors), for instance, of 0.6 to 0.8 °C for mean monthly air temperatures, of 9 to 12 mm for mean monthly precipitation sums and of 2 to 15 cm for mean snow depth at mid-month. Interpolation quality is thus mostly of a similar magnitude as measuring accuracy.

 

 

 

 

References:

Daly C, Halbleib M, Smith JI, Gibson WP, Doggett MK, Taylor GH, Curtis J, Pasteris PP (2008): Physiographically sensitive mapping of climatological temperature and precipitation across the conterminous United States. Int J Climatol 28:2031–2064. doi: 10.1002/joc.1688

Daly C, Neilson RP, Phillips DL (1994): A statistical-topographical model for mapping climatological precipitation over mountainous terrain. J Appl Meteorol 33:140–158. doi: 10.1175/1520-0450(1994)033<0140:ASTMFM>2.0.CO;2

Frei C (2014): Interpolation of temperature in a mountainous region using nonlinear profiles and non-Euclidean distances. Int J Climatol 34:1585–1605. doi: 10.1002/joc.3786

Frei C, Schär C (1998): A precipitation climatology of the Alps from high-resolution rain-gauge observations. Int J Climatol 18:873–900. doi: 10.1002/(SICI)1097-0088(19980630)18:8<873::AID-JOC255>3.0.CO;2-9

Hiebl J, Frei C (2014): Daily temperature grids for Austria since 1961 – concept, creation and applicability. Theor Appl Climatol (submitted)

Olefs M (2013): Projekt APOLIS – Austrian photovoltaic information system. Final report. Vienna: Zentralanstalt für Meteorologie und Geodynamik, 70 pp.

Olefs M, Schöner W (2012): A new solar radiation model for research and applications in Austria. In: EGU General Assembly 2012, Vienna, 22.–27.04.2012

 

Joomla SEF URLs by Artio