(4) If the plant cover increases, the intensity of radiation in the shortwave infrared band decreases non- 
linearly. 
(5) The progress of spectral signatures of different soil types is relatively monotonous. Low-grade re 
mission can be expected in the bands near 0.4 and 2.5 urn with high concentrations of humus and water 
in the surface layer. Absorption bands in the shortwave infrared bands are caused by high water content 
/fig. 1.2/. 
(6) Thermal emission of wet soil will be determined by transpiration cooling. If the soil consists of the 
same material it will be determined by the size of the surface and its structure 
Analysis of the spectral signature may be complicated by any disturbing factors, for example varying 
water content or the presence of similar absorption characteristics in different materials. In such situa 
tions the formation of quotients for the quantities of radiation in different wavelength bands can be of 
assistance. To determine the vitality of vegetation, for example: 
I0.66/I0.8 rsp. (lo.8-Io.66)/(Io.8-»-l0.66) 
and to determine Alumosilicates in geological analysis: 
I2.20/I2.35 rsp. I2.22/I2.35 
Research and simulation tests have demonstrated that complex analysis of multisensor data is a pro 
mising way to resolve many problems in collecting ecological information. The criteria for evaluation of 
the conditions of vegetation and soil by complex analysis of remotely sensed data in the VNIR-, SWIR- 
, TIR- and MW- wavelength bands are given in fig.2. 
In order to meet the varying demands of different scientific tasks in connection with the geometric and 
spectral resolution of remote sensing data, it is necessary to form some kind of system. Fig. 3 shows 
an attempt to do this. 
Airborne remote sensing systems are able to map out high resolution (geometric and spectral) data, but 
they cannot cover large areas of the earth’s surface.For this reason it is very important in solving many 
problems of ecology to analyse spaceborne and airborne remote sensing data as well as ground-based 
observations. 
An other question is the problem of the rate and quantity of data, because we are not able to record it 
to an unlimited extent. If a medium geometric resolution of multispectral data is decided on, attention 
must be paid to the mixed pixels problem. In this situation it may be helpful to obtain additional high 
geometric resolution data in panchromatic form. The concept TOMAS includes such a system. It should 
be possible to analyse the panchromatic data set of a scene in order to determine the homogenety of 
the radiation intensity of multispectral pixels./fig. 4/.