Comparison of observed and model simulated data
Verification of the numerical global climate model (GCM) is the essential step
for ascertainment application of model data for description of the actual climate
and water balance parameters at present and for estimation of their possible
changes during the period in consideration. The ECHAM-4 model data designed
by Max-Planck Institute for Meteorology were used (Bengtsson 1997). The grid
resolution of this GCM is 28°x28° in latitude and longitude, the vertical domains
extend up to 30 km height with time step of one month. The set of grid points
with regard to the studied area is shown in Fig.1. Simulated data (about 40
parameters, including air temperature, total precipitation, and others) were
obtained for the 1960-1999 period and for two scenarios for the 2000-2050 period
-
a) the case of doubling of greenhouse gases (hereinafter we denote it as G-scenario),
and
b) the case of doubling of greenhouse gases with taking into account direct
effect of atmospheric aerosols (GA-scenario).
For calculation mean as a whole annual air temperature and precipitation the multiple regression method was used. The next equations for the territory of Karelia were obtained,
T=0.346+0.91T17-0.75T13-0.35T7, (6)
P=132+2.27P12 +0.32 P17 –1.7P13, (7)
where T and P are annual air temperature and precipitation for the territory, respectively, Ti, and Pi are annual air temperature and precipitation in the corresponding grid points (model data in accordance with their numbers in Fig 14).
Fig. 14. Model (1) and observed (2) annual air temperature for the territory of North-West of Russia a for the control period 1960-1999
To estimate conformity of monthly averaged observed annual air temperature and precipitation and model data obtained on the separate weather stations, they were compared with those in the corresponding grid points, situated nearby. Analysis of the superposed observed and model data for separate weather stations showed a good correspondence in monthly and annual air temperature (Fig. 15), but it does not satisfactory fit for monthly precipitation.
References can be found here.