Abstract. We developed a mobile remote sensing measurement facility for spectral and anisotropic reflectance measurements. We measured reflection properties (BRF) of over 100 samples from most common land cover types in boreal and subarctic regions. This extensive data set serves as a unique reference opportunity for developing interpretation algorithms for remotely sensed materials as well as for modelling climatic effects in the boreal and subarctic zones.

Our goniometric measurements show that the reflectances of the most common land cover types in the boreal and subarctic region can differ from each other by a factor of 100. Some types are strong forward scatterers, some backward scatterers, some reflect specularly, some have strong colours, some are bright in visual, some in infrared. We noted that spatial variations in reflectance, even among the same type of vegetation, can be well over 20%, diurnal variations of the same order and seasonal variation often over a factor of 10. This has significant consequences on the interpretation of satellite and airborne images and on the development of radiation regime models in both optical remote sensing and climate change research.

We propose that the accuracy of optical remote sensing can be improved by an order of magnitude, if better physical reflectance models can be introduced. Further improvements can be reached by more optimised design of sensors and orbits/flight lines, by the effective combining of several data sources and better processing of atmospheric effects. We conclude that more extensive and systematic laboratory experiments and field measurements are needed, with more modelling effort.