de Jong, Steven 
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
  
Surface temperatures 
Bare Soil 
* 
o Grassland 
8 
2 Garrigue 
= Vineyard 
5 zz DRST: 
E oH Maquis 
E 
TTT Pine forest 
0 30 60 90 120 150 180 210 240 270 
Time steps [10 min] 
  
  
  
Figure 4: Resulting surface temperature values of the calibrated model, for each landcover type 
The regional distribution of the surface temperatures is mainly the result of differences in landcover. Bare soils have 
higher surface temperatures then vegetated areas. An other factor controlling the regional temperature distribution, is 
the slope aspect. Slopes facing the north, receive less sunlight, and are consequently colder (figure 5). 
  
   
   
   
     
       
Liat ix 
CEA 
E d 
5 CL 
> 
E 0r] 
ay 
EI 
RI 
LE 
d 
SES 
ER 
[4 E. 
Hd 
HEA 
i 
1i 
EET 
   
       
       
    
  
   
          
  
  
   
   
   
   
m ud 
AE MEM Ea 
Legend | ) - dn 
Temperature in ee E A e E 
            
       
  
© 
ij 
i 
3 
dearees Celsius 
TR 
jh 
pe 
a 
i 
) 
x3 
e 
SO 
AS 
   
  
  
  
i 
d 
at 
d 
n 
ena 
nee 
M 
ca 
paye ENS SN 
aa OA 
ER EE par 
     
  
  
  
Figure 5: calibrated model surface temperature result of the study area at 29-06-1998 at 12:00 
6 DISCUSSION AND CONCLUSIONS 
In this study optical and thermal images acquired by DAIS during the Peyne experiment are integrated and used to run, 
validate and calibrate a regional simulation model of surface temperature. Additional model input and validation 
information is derived from field surveys and from literature. DAIS image analysis shows that image quality was 
acceptable in visible and near infrared wavelengths but that striping in the shortwave infrared bands hampers the use of 
this spectral part. Empirical line methods were successfully applied to convert radiance data of the optical and thermal 
bands into reflectance and surface temperature respectively. 
All surveyed landcover types differ significantly of each other regarding their surface temperatures. From the model and 
survey results, it is concluded that vegetation cover is the major factor controlling surface temperature. Interpolation of 
the soil properties (soil moisture content, porosity and bulk density) is achieved by applying geostatistical interpolation 
techniques. The computed variograms of field data are of satisfactory quality. The interpolation technique used is 
conditional simulation and the resulting maps give insight in the spatial distribution of data uncertainty. The uncertainty 
of the interpolated maps are within an acceptable range for running the spatial version of the model. The three resulting 
maps are used as input in the surface temperature model. 
The model used for predicting surface temperature is a physically based model using the surface energy balance for 
calculating surface temperatures at a regional scale. The model collects input variables from the DAIS imagery, 
meteorological data, field measurements and literature. The model is most sensitive to land use and the resistance 
values, used in calculation of the latent heat flux. Preliminary model results show that the model gives a fairly good 
average estimate of a 48 hour surface temperature cycle of the study area. 
  
International Archives of Photogrammetry and Remote Sensing. Vol. XXXIII, Part B7. Amsterdam 2000. 353