Airborne SAR Acquisitions in Northern Finland in Support of the COREH2O Mission
Lemmetyinen, Juha1; Pulliainen, Jouni1; Kontu, Anna1; Meta, Adriano2; Coccia, Alexander2; Schneebeli, Martin3; Proksch, Martin3; Derksen, Chris4; Schüttemeyer, Dirk5; Kern, Michael5; Davidson, Malcolm5; Rott, Helmut6; Voglmeier, Karl6
1Finnish Meteorological Institute, FINLAND; 2Metasensing, NETHERLANDS; 3WSL-SLF, SWITZERLAND; 4Environment Canada, CANADA; 5ESA, NETHERLANDS; 6ENVEO IT GmbH, AUSTRIA
The proposed CoReH2O (Cold Regions Hydrology High-Resolution Observatory) mission, one of the candidates for the 7th Earth Explorer Core mission by the European Space Agency (ESA), aims to provide information on Snow Water Equivalent (SWE) over land at an unprecedented spatial resolution. In recent years, numerous experimental campaigns have been conducted in support of the mission scientific goals. In the winter season of 2011-2012, ESA launched an experimental campaign in Northern Finland using the SnowSAR instrument, an airborne SAR designed to operate on the same X- and Ku frequency bands as CoReH2O. The campaign follows a successful test flight in the same region in March 2011. Three test sites were chosen for the acquisitions; the main test site at the Finnish Meteorological Institute Arctic Research Centre (FMI-ARC), near the town of Sodankylä in Northern Finland, represents a typical boreal forest/taiga environment. The landscape is dominated by coniferous forests, lakes and wetlands (bogs). Snow cover typically persists for five months during winter; the measured 30 -year average of maximum SWE in the region is over 180 mm. The main test site is also the location of the ESA NoSREx (Nordic Snow Radar Experiment) campaign, a three-year effort to measure X- to Ku band backscatter of snow cover from a fixed location using the ESA SnowScat scatterometer. The presence of the ground-based scatterometer, as well as an extensive archive of space-borne TerraSAR-X (X-band) acquisitions from the site, offers good opportunities for instrument cross-calibration. A second test site over land was chosen from the Saariselkä tundra region, 150 km north from the main test site. The general topography was varying with several low-lying tundra hills. In contrast to the main test site, the region represents a sparsely vegetated and rocky tundra landscape, with significant variability in snow distribution due to topography and wind drift effects, making the site ideal for high-resolution airborne acquisitions. A third test site was chosen over sea ice; the aim was to provide first images of X- to Ku band backscatter of snow cover on sea ice, a secondary mission goal of CoReH2O. TerraSAR-X acquisitions were also made from these sites as a reference to the airborne observations.
Over ten data acquisition flights were conducted at the main test site; each airborne mission attempted to cover the same area using a mosaic of up to 30 flight transects, each ca. 7 km long. Acquisitions were timed to correspond closely to the planned CoreH2O revisit times during two phases of the mission (a 3 and 15-day revisit time), thus demonstrating the mission concept both spatially and in time. The acquisitions represent a range of snow conditions, from early season shallow fresh snow to a late-winter snow pack condensed by several melt-refreeze events. The last airborne acquisitions took place close to the maximum snow conditions of the season. Results indicate that the influence of increasing snow cover can be detected in the X and Ku band radar signatures. Retrieval experiments from the airborne acquisitions with the baseline CoReH2O retrieval algorithm match in situ observations of snow conditions with good accuracy over non-vegetated areas. For densely vegetated areas, the forest canopy masks out the signal from snow cover and the retrieval cannot be performed.
The tundra site was visited twice, first in early winter and for a second time in mid-winter conditions. Acquisitions indicate a high spatial variability in snow cover especially during the last acquisition, with an over 100 mm overall increase in SWE measured between the two acquisitions. A single acquisition was performed over the sea ice site; although no retrieval attempts of SWE over sea ice were made, we show first results of the X- and Ku band backscatter signal variations with snow cover properties over the ice.