Modeling primary production using a 1 km daily meteorological data set

Abstract

The availability of daily meteorological data extended over wide areas is a common requirement for modeling vegetation processes on regional scales. The present paper investigates the applicability of a pan-European data set of daily minimum and maximum temperatures and precipitation, E-OBS, to drive models of ecosystem processes over Italy. Daily meteorological data from a 10 yr period (2000 to 2009) were first downscaled to 1 km spatial resolution by applying locally calibrated regressions to a digital elevation model. The original and downscaled E-OBS maps were compared with meteorological data collected at 10 ground stations representative of different eco-climatic conditions. Additional tests were performed for the same sites to evaluate the effects of driving a model of vegetation processes, BIOME-BGC, with measured and estimated weather data. The tests were carried out using 10 BIOME-BGC versions characteristic for local vegetation types (Holm oak, other oaks, chestnut, beech, plain/hilly conifers, mountain conifers, Mediterranean macchia, olive trees, and C3 and C4 grasses). The experimental results indicate that the applied downscaling performs best for maximum temperatures, which is the most decisive factor for driving BIOME-BGC simulation of vegetation production. The downscaled data set is particularly suitable for the modeling of forest ecosystem processes, which could be further improved by the use of information obtained from remote sensing imagery.

Correction of a 1 km daily rainfall dataset for modelling forest ecosystem processes in Italy

“…The application of the downscaling procedure proposed by Maselli et al () to the 30 year E‐OBS dataset provided 1 km estimates whose main properties are fully discussed in the present study. The mean annual rainfall map obtained by averaging these estimates is shown in Figure .…”

“…In particular, a pan-European land with 0.25 ∘ resolution gridded dataset, E-OBS, has been released recently for daily precipitation, sea-level pressure and minimum and maximum temperatures (Haylock et al, 2008;van den Besselaar et al, 2011). Maselli et al (2012) proposed a method to downscale this dataset to 1 km spatial resolution using spatially weighted regressions. The method was applied in Italy over a period of 10 years (2000)(2001)(2002)(2003)(2004)(2005)(2006)(2007)(2008)(2009), producing a consistent daily dataset whose main limitation is a marked and variable tendency to under-estimate rainfall (Maselli et al, 2012).…”

“…Maselli et al (2012) proposed a method to downscale this dataset to 1 km spatial resolution using spatially weighted regressions. The method was applied in Italy over a period of 10 years (2000)(2001)(2002)(2003)(2004)(2005)(2006)(2007)(2008)(2009), producing a consistent daily dataset whose main limitation is a marked and variable tendency to under-estimate rainfall (Maselli et al, 2012).…”

Use of geographically weighted regression to enhance the spatial features of forest attribute maps

“…36 Similarly, GWR procedures can be adapted to cope with data layers having different spatial resolutions. 38 The application of GWR to the latter case assumes that a significant local correlation exists between the low spatial resolution predictions of vegetation properties and remotely sensed data. 38 The application of GWR to the latter case assumes that a significant local correlation exists between the low spatial resolution predictions of vegetation properties and remotely sensed data.…”

Spatial downscaling of European climate data

“…In Europe, scientists are forced to focus either on individual countries or regions using national data sets (Hasenauer et al, 2003;Venäläinen et al, 2005;Maselli et al, 2012). Another common way to study climate on larger scales in Europe is to use point data, from weather stations or flux towers, as representive of larger areas (Janssens et al, 2001;Ciais et al, 2005).…”