The L2 products from the Maja chain are regularly evaluated for their quality, whether in terms of cloud detection, optical thickness estimation or surface reflectance estimation. While the validation of the cloud masks is essentially based on the tedious creation of sets of controls (manual or semi-automatic), the estimation of the optical properties of the atmosphere can be based on ground measurements performed with a photometer.
Photometers used in environmental science allow directional measurement of direct and diffuse solar radiation in the visible to near-infrared range, with a usual spectral sampling of 9 to 12 bands. Properties of the atmosphere such as optical thickness, aerosol volume size distribution, refractive index and water vapour content can be deduced from these measurements. This is precisely the purpose of the AERONET network (AErosol RObotic NETwork) which is based on the instrumentation developed by the French company CIMEL. It is thus possible to validate the optical thicknesses derived from Maja directly from the measurement.
To go even further, the CNES developed in 1997 a protocol using a photometer to carry out not only angular and spectral measurements of the incident radiation, but also of the radiation reflected by the surface. It is thus possible, after processing, to deduce the bi-directional reflectance (BRDF) of the surface of the measurement site, and thus validate the MAJA surface reflectance products. The first site deployed by CNES in 1997 is located in La Crau (South of France), and was completed in 2017 by another ROSAS station in Gobabeb (Namibia), in cooperation with ESA and NPL. Both sites have an extremely homogeneous land cover, with pebbles and short grass at La Crau, and arid desert sand at Gobabeb, and both intra- and inter-annual variability in surface condition remains low. They are therefore ideally suited for instrument calibration and validation of L2 products.
However, the two sites of La Crau and Gobabeb have the disadvantage of limiting validation to rather ideal cases. In practice, we should be able to evaluate the performance of Maja on vegetated canopies, and in particular on heterogeneous areas, since the chain includes the correction of adjacency effects. Our project is therefore to install a ROSAS station on the agricultural area of the Lamothe farm in Lamasquère, a CESBIO perennial measurement site maintained within the framework of the South-West Regional Space Observatory. Located a few kilometres South-West of Toulouse, this plot of land of about 24 ha follows a crop rotation alternating maize, winter wheat and intermediate crops. It is bordered to the East and to the West by wooded areas, which promise beautiful adjacency effects.
Following the CNES ROSAS system specifications, the CESBIO is now equipped with a brand new CIMEL CE318-TU12 photometer which will be installed at the top of a 10-meter mast on the site indicated by the red and yellow dot on the map. Equipped with its own datalogger and solar panels, the station will transmit the measurements directly to CNES by 4G. Data processing and distribution will be carried out by CNES within the framework of ROSAS.
At this stage (March 2020), we have received the instrumentation as well as the mast, just before the containment due to the COVID-19 virus. The installation of the mast will take place after the winter wheat harvest, probably around next August after the current wheat crop. This will be followed by the installation of the CIMEL instrument and the testing of the whole chain, with the first sets of data available at the end of the year. The data should then be inserted into the RadCalNet network, but installing such a facility in an agricultural plot is not a bed of roses. We will have to cope with the cultural cycle and calendar, take into account the mechanized interventions, and the irrigation, every second year.
To be continued!