ANALYSIS OF AN ERS-1 SAR TIME SERIES AND OPTICAL SATELLITE DATA FOR FORESTRY APPLICATIONS IN
TEMPERATE ZONES
G. Kattenborn”, E. Nezry”*
*University Freiburg, Dep. Remote Sensing and Landscape Information Systems, Germany
E-mail: kattborn@combo.forst.uni-freiburg.de
**PRIVATEERS Ltd., St. Maarten, Dutch Antillas
E-mail: edmond.nezry@iol.it, Phone (++39)332 781494
Commission VII, Working Group 2
KEY WORDS: Forestry, vegetation, inventory, monitoring, multitemporal, SAR, SPOT, fusion
ABSTRACT:
Due to their weather independent operation there is a growing interest in utilising imagery collected by imaging synthetic aperture ra-
dars on satellites like ERS-1, JERS-1 and RADARSAT. Their data, acquired on a regular basis, at short intervals, over the entire
planet, will allow a variety of forest ecosystem characteristics to be monitored for the years to come. To take full benefit of these advan-
tages it is now necessary to investigate the information content of these data under a variety of scene related and environmental condi-
tions. The objectives of this study are to exploit multidate ERS-1 observations of the Freiburg-Black Forest test-site in Germany and a
rich available ground truth data set by developing techniques for change detection and monitoring in order to assess the information
content of ERS-1 SAR data for forestry and other land applications, and especially to evaluate the potential for improved information
extraction from multitemporal SAR data. Furthermore the combined use of optical and microwave spaceborne data should give insight
into possible synergetic effects for qualitative and quantitative analysis of forest stands. Changes in backscatter observed by the ERS-1
SAR have been related to the variation of meteorological parameters for different landuse. It confirms that the temporal sequence of the
C-band reflectivity of vegetated areas measured by ERS-1 appear to be sensitive to changes in moisture conditions and effects of tem-
perature on dielectric properties. Long term variations due to phenological and seasonal effects can then be used to discriminate land
use classes. Temporal backscatter profiles show that a good discrimination of vegetated areas (grassland, agriculture, forest) is
achieved from October to May. Further results from this investigation demonstrate that the C-VV band of ERS-1 is sensitive to age
(and therefore closely related parameters such as tree height, woody biomass) in young deciduous and coniferous forest stands. A syn-
ergetic relationship between optical and microwave data has been established. The combined data set allowed better differentiation of
forest age classes, as compared to results obtained with the individual sensors. The results of this study add to the growing evidence of
the valuable capabilities imaging radars have for the studying and monitoring of forests due to the demonstrated sensitivity to a variety
of processes in vegetation ecosystems and the proven synergy to optical satellite data.
1. INTRODUCTION backscatter. Especially important for the use of SAR for forest
analysis is the sensitivity of radar backscatter to the water status
Due to their weather independent operation there is a growing of trees and understory, which can be affected by seasonal
interest in utilizing imagery collected by imaging synthetic aper- growth or senescence, and by external conditions such as insola-
ture radars (SAR), with the launch of the first European Space tion, water stress, temperature, floods, snow, rain, freeze, thaw,
Agency (ESA) Earth Remote Sensing Satellite (ERS-1) in July etc., as is clearly stated by theoretical models, ground measure-
1991, the first Japanese Earth Resources Satellite (JERS-1) in ments and previous remote sensing campaigns. The ability of mi-
February 1992, and the Canadian RADARSAT in 1995. Their ^ crowaves to penetrate the outer part of forest canopies and be
data, acquired and archived on a regular basis, at short intervals, scattered by branches and trunks leads to the assumption that
over the entire planet, will allow a variety of forest ecosystem SAR sensors might be able to provide estimates of forest bio-
characteristics to be monitored for the years to come. Compared ~~ mass or other stand attributes for both ecological investigations
to the multispectral capabilities of optical spaceborne sensors the of carbon storage and as an economical source of forest man-
information content of a single frequency SAR, such as the ERS- agement information. However, the success of any monitoring
| AMI may be limited for land and forestry applications. How- approach in observing longer term changes in forest attributes
ever, continuous spaceborne SAR coverage may provide more in- such as biomass level or successional stage depends on a sound
formation than can be obtained with single or limited optical ^ understanding of the masking effects of meteorological, phe-
spaceborne observations. ERS-1 SAR data will actually allow nological and seasonal changes.
eventual temporal gaps of optical coverage to be filled, as well as
to ensure radar cost-efficient coverage of the site over the year. From that, The objectives of this study are to exploit multidate
To take full benefit of these advantages it is now necessary toin- ~~ ERS-1 observations and a rich available ground truth data set in
vestigate the information content in ERS-1 SAR data under a order to assess the information content of ERS-1 SAR data for
variety of scene related and environmental conditions. forestry and other land applications, and especially to evaluate
, the potential for improved information extraction from mul-
" the particular case of forests, the seasonal growing cycle re- titemporal SAR data. Furthermore the combined use of optical
sulting in leaf-on/leaf-off conditions, budding, cone emergence, and microwave spaceborne data should give insight into possible
caf growth, branch growth, change of forest floor in composi- synergetic effects for qualitative and quantitative analysis of for-
tion, and understory cover, is expected to influence SAR est stands.
331
International Archives of Photogrammetry and Remote Sensing. Vol. XXXI, Part B7. Vienna 1996
