RAIN 
iration 
NS how the 
| and point 
regression 
1986) and 
isted of a 
?0, 1986 at 
dinavia. All 
short wave 
990, 1992 
land use 
1988) from 
>AR-albedo 
Radiation, 
irradiation 
ort Wave 
alibrated by 
ric counter 
radiometric 
formed by 
available to 
| calibrated 
ver the sea 
on formula 
5tt/Volchok, 
Yated using 
of J. Price, 
, Scherer, 
teorological 
area, were 
teorological 
A profile of 
been taken 
of satellite 
  
4. STUDY AREA 
The Abisko National Park is situated in northern Sweden 
in the province Lappland at 6821'N/1999'E. The study 
area in the sub province Torne Lappmark includes the 
northern part of the Abisko National Park, part of lake 
Abisko, the hills south of Abisko Ostra and the crests of 
mount Njulla (1169m) and Slattatjakka (1191m). In 
Scandinavia the northern limit of boreal forest is formed 
by the birch. The forest limit in the study area is 
determined by elevation and situated between 550m and 
650m. It can be differentiated between the "subalpine 
birch tree region" and the "alpine region" of alpine heaths, 
meadows and tundra. 
5. THE SLOPE WIND MODEL OF BREHM 
Under stable conditions, the atmospheric boundary layer 
shows special characteristics at slopes, and a sun driven 
mecanism forces the air into a well defined movement 
upwards. 
Brehm (Brehm, 1986) parametrized the slope wind by 
atmospheric stability, excess temperature (temperature 
difference between the terrain at roughness length-height 
and free atmosphere) and slope angle. 
His model generates values for the geostrophic friction 
coefficient c4, the heat transfer coefficient Cp and their 
ratio KR Cg/Ch: which are defined by the following 
equations: 
  
^ (1) 
K u* Ox / 
n ah (2) 
Where u* is the friction velocity, 6* the characteristic 
temperature fluctuation, « Von Karman's constant, A the 
excess temperature, p the coefficient between the 
acceleration due to gravity and the potential temperature 
of the surface g/©, and the temperature gradient of the 
free atmosphere d@/dz. 
BREHM generated values for c, and n in dependance of 
the analog soil rosby number Ro for different slope 
angles. Soil rosby number in turn is defined as follows: 
- KA 3 
Y Sina, zo (3) 
Zo is the roughness length. 
In order to calulate the analog soil rosby number for each 
terrain pixel, we need the surface temperature, a vertical 
temperature profile of the free atmosphere, a method to 
Parametrize roughness length and a digital terrain for 
slope angles and elevations. 
The relations between rosby number and c, or m, 
calculated by Brehm's slope wind model, are graphically 
"presented in Brehm's study (Brehm, 1986) in two 
graphics for different slope angles. The functional 
dependance between cg / n and the rosby number for 
661 
any slope angle were fitted by a polynomial equation, so 
that c, and n could be written as functions Cg = f(a,Ro) 
and n = f(a,Ro). 
6. PARAMETRIZATION OF SENSIBLE HEAT FLUX 
Sensible heat flux can be written as follows: 
H=-p cpus Ox (4) 
where p is the air density and Cp is the specific heat of air 
at constant pressure. 
Replacing the parameters with de definitions of c, and n 
in formula (1) and (2), equation (4) can be reformulated 
in the following way: 
Hzp Cp «cg (Roa) n(Ro,a) ABH (5) 
The excess temperature as formulated by Brehm is the 
temperature difference between the free atmosphere and 
the surface defined at a level above the ground that 
corresponds to roughness length. 
Mannstein (Mannstein, 1991) assumes, that the radiation 
temperature derived from thermal sensors of a remote 
sensing devices represent the skin temperature of the 
ground which is best represented at zero height. 
Therefore, he adopts a formulation of Monin and 
Zilitinkevich (Monin/Zilitinkevich, 1968) which expresses 
the difference Ay between surface and zj-level as 
follows: 
Ag = -0.13 ©+ (tapas (6) 
v is the kinematic viscosity of air. 
Using the definitions for Cg and n, Aq can be written as: 
s: : eg € AN BAT za 0.45 
Ag 7 -0.13« n cg A( x ) (7) 
For a given temperature difference A, between the 
ground (derived by thermal sensors) and free 
atmosphere, the temperature difference A between the z, 
level and the atmosphere, as it will be used for the 
model, can be found by itereation, so that the following 
equation holds: 
Ag = ^ * Aq (8) 
As is the known temperature difference between satellite 
temperature at n=0 and free atmosphere. 
7. DETERMINATION OF THE MODEL PARAMETERS 
7.1 Modelling of roughness length 
Roughness length varies widly within a given land use 
class. Mannstein distinguished only 3 land use types and 
substituted each of them with a single value for 
roughness length. 
International Archives of Photogrammetry and Remote Sensing. Vol. XXXI, Part B7. Vienna 1996