NORA and GlobMOS – Retrieval of Soil Moisture from L-band Radiometer Data
Soil moisture stored in surface soils is an essential element of the water cycle regulating water and energy exchanges between the land and lower atmosphere, affecting plant growth and crop yield, and regulating carbon dioxide exchanges from and to the atmosphere.
Satellite remote sensing can produce global information on these variables provided suitable sensors and data interpretation methods are available. The most promising method for global soil moisture and ocean salinity monitoring is space-borne microwave radiometer measurements at a low microwave frequency.
The ESA SMOS satellite, launched in November 2009, produces global data on these parameters. SMOS is the first space-borne 2-dimensional (2-D) interferometric microwave radiometer for Earth Observation. It operates in the 1.400-1.427 GHz band reserved for radio astronomy and thus suitable for the space-borne radiometers as well.
The goal of the NORA project, started in 2006 and funded by Tekes, is to improve the stability of the SMOS sensor and the quality of SMOS geophysical retrieval products, especially for northern areas. The main objective of the GlobSMOS project, started in 2011 and funded by the Academy of Finland, is to develop advanced methods for soil moisture and ocean salinity retrieval from L-band radiometer data.
The projects employ satellite data from SMOS, airborne data from Aalto University’s HUT-2D L-band interferometric radiometer (completed in 2006; similar to the SMOS instrument); accommodated onboard Aalto University’s research aircraft), and relevant in situ data. Also the 6.8 and 10.7 channels of the Aalto HUTRAD radiometer are used as an auxiliary data source.
Figure 1. Aalto University’s airborne radiometers on the Short Skyvan research aircraft. The interferometric L-band HUT-2D is mounted beneath the aircraft fuselage and the multifrequency HUTRAD radiometer is in the rear cargo area (door removed).
The HUT-2D interferometric instrument is a unique airborne sensor producing datasets with the characteristics of the data for algorithm development and validation close to those acquired by the SMOS sensor. The high spatial resolution of HUT-2D is an important asset for detailed studies.
The resolution of the final SMOS brightness temperature grid is about 50 km x 50 km. Forests attenuate microwave emission from the ground surface, thus masking the effect of soil moisture to the brightness temperature. The mixed pixel problem is due to the effect of various land-use categories (forest, agricultural land, swamps, water) within each SMOS pixel to the observed brightness temperature. These two effects may cause substantial errors to the retrieved soil moisture values; thus, they have to be taken into account as SMOS data are applied to continental regions such as northern parts of Eurasia and North America.
High-resolution airborne L-band data for selected test sites are combined with vegetation/land-use data (e.g. CLC2000) and ground reference data in order to experimentally determine the behaviour of the brightness temperature for typical boreal forest, and the response to various land-use categories. Based on information concerning land use and forest within a SMOS pixel, the algorithm to be developed computes the soil moisture value from the observed brightness temperature. The final algorithm is verified with surface reference data and data from the SMOS satellite.
Specifically, we aim to:
• Study vegetation parameterization in the Finnish boreal forest environment and evaluate the suitability of existing soil moisture retrieval models and algorithms to Finnish environment using radiometer measurements.
• Based on the experimental data sets, develop a new emission model for the typical Finnish boreal forests and wetlands to be used with the SMOS soil moisture retrieval algorithm.
The test sites are located in Rajamäki, near Helsinki in southern Finland, and in Sodankylä, northern Finland.
Figure 2. A soil moisture map based on HUT-2D measurements. Soil moisture color coded.
Figure 3. In-situ measurements of soil moisture and vegetation parameters in Sodankylä, northern Finland, in 2009.
Cooperation: Finnish Meteorological Institute, Finnish Environment Institute