Compensation of Forest Effect for the Retrieval of Snow Mass: A Contribution to the CoReH2O Mission
Macelloni, Giovanni1; Montomoli, Francesco1; Brogioni, Marco1; Fontanelli, Giacomo1; Lemmetyinen, Juha2; Pulliainen, Jouni2; Karl, Voglmeier3; Hajnsek, Irena4; Kern, Michael5; Rott, Helmut3
1IFAC - CNR, ITALY; 2FMI, FINLAND; 3Enveo, AUSTRIA; 4DLR- ETH, GERMANY; 5ESA - ESTEC, NETHERLANDS

The COld REgions Hydrology High-resolution Observatory (CoReH2O) is one of the three satellite missions selected for scientific and technical feasibility studies (Phase-A) within the Earth Explorer Program of the European Space Agency [1]. The principal objective of the CoReH2O mission is to carry out the frequent spatially-detailed measurements of snow and ice in order to advance our knowledge and prediction of the water cycle in cold regions. The sensor proposed is a dual frequency SAR, operating at Ku-band (17.2 GHz) and X-band (9.6 GHz), VV and VH polarizations, with a swath width of about 100 km. The principal products obtained from the mission will be the estimations of the extent, the snow water equivalent (SWE) and melting state of the seasonal snow cover and the snow accumulation on glaciers. The presence of vegetation has a significant impact on the propagation of the radar signal at X- and Ku-band, depending on its structure, biomass, water content and cover fraction. In particular for dense forest, the scattering of vegetation strongly hides the signal from snow and, consequently, compromises the sensitivity to snow parameters [2]. During Phase-A of the mission an approach for correcting the effect of forest have been developed and implemented. The proposed approach is based on the introduction in the retrieval algorithm of a semi-empirical model able to simulate the backscattering from a forest stand. The approach has been developed for boreal forest which represents the most typical forest stand in the region of interest for the mission. The e.m. model is a simplified version of a more complete radiative-transfer model, and uses as inputs auxiliary data like vegetation type, cover fraction and tree height. In order to evaluate the response of the developed model to the variability of the inputs, a sensitivity analysis was conducted as a function of SWE and forest parameters. In particular, the sensitivity of 0 to SWE was conducted for different biomass values in order to evaluate the maximum biomass (or cover fraction) acceptable for retrieval purposes. Simulations were carried out at both X- and Ku-band at 40° of incidence angle, with an SWE ranging from 0 to 300 mm and varying the cover fraction of forest in the 0-0.4 range. The analysis pointed out, as expected, that vegetation strongly influences the sensitivity of backscattering to snow parameters: in particular, by increasing cover fraction (i.e. cover fraction) the sensitivity to SWE decreases until it becomes negligible when cover fraction is higher than 0.3. From an operational point of view, this means that areas showing these values must be masked out on the retrieval process line of the mission. The model was validated by using both airborne and satellite SAR data acquired over the Sodankylä test site in Finland, where several campaign were carried out in recent years within the framework of mission preparation activities. Finally the complete SWE retrieval algorithm was tested using the data acquired over the same area and the developed forest model. The comparison between measured and retrieved SWE values confirms that using proposed approach is possible to meet the CoReH2O mission requirements in sparse forested area. Possible future developments and improvements will be also presented.

Please note that This is a placeholder abstract from EE7 related science preparatory activities which is relevant to the Cryosphere science theme, or Future EE Session talk (pending EE7 mission selection in March)


REFERENCES

[1] Rott H., Yueh et al., "Cold Regions Hydrology High-Resolution Observatory for Snow and Cold Land Process". IEEE Proceedings, 98 (5): 752-765, 2010.

[2] G.Macelloni, M.Brogioni, F.Montomoli, G.Fontanelli: "Effect of forests on the retrieval of snow parameters from backscatter measurements", European Journal of Remote Sensing , 45: 121-132, 2012.