NRT Product available since Mar 2018
The gross primary production (GPP), i.e. the rate at which vegetation converts light into chemical energy by photosynthesis, is an essential parameter to characterize the ecosystem processes. The assessment of GPP on wide areas, which is necessary to study the global carbon cycle and for planning and managing resources in response to changing environmental conditions, can be performed using procedures driven by remote sensing data.
This operational product is documented in the following documents:
Please see Product Peer-Review publications in References.
The use of LSA SAF products in publications is kindly requested to be duly acknowledged:
MGPP was provided by the EUMETSAT Satellite Application Facility on Land Surface Analysis (LSA SAF; Trigo et al., 2011)
Trigo, I. F., C. C. DaCamara, P. Viterbo, J.-L. Roujean, F. Olesen, C. Barroso, F. Camacho-de Coca, D. Carrer, S. C. Freitas, J. García-Haro, B. Geiger, F. Gellens-Meulenberghs, N. Ghilain, J. Meliá, L. Pessanha, N. Siljamo, and A. Arboleda, 2011: The Satellite Application Facility on Land Surface Analysis. Int. J. Remote Sens., 32, 2725-2744, doi: 10.1080/01431161003743199
The GPP product from MSG/SEVIRI data (MGPP) is currently generated at 10-day temporal resolution for the full spatial resolution of the MSG/SERVIRI instrument. A daily GPP product (MDGPP) is also generated internally to compute the 10-day GPP composite. The MGPP product is based on an ensemble of operational LSA SAF products, such as the downward surface shortwave radiation flux (DIDSSF) (LSA‐203) and the daily fAPAR (MDFAPAR) (LSA‐425). The actual evapotranspiration (ET) (DMET, LSA-302) and the reference ET (DMETREF, LSA-303) products are also considered to account for the reduction in light use efficiency (LUE) due to water stress. The global land cover (GLC2000) is used to discriminate the main land covers presented in the full MSG disk. The GPP is not computed for desert areas due to the lack of values for some inputs (e.g. DMET).
The GPP is expressed in the range from 0 to 15 g m-2 day-1. It provides a layer with uncertainty information in the estimates, which ranges from 0 to 10 g m-2 day-1. The product performance is assessed by means of direct comparison of GPP estimates with in-situ eddy covariance (EC) tower GPP estimations. Moreover, the consistency against other synergist global remote sensing products, such as MODIS GPP product (MOD17A2) and FLUXCOM GPP product derived at the Max Planck Institute for Biogeochemistry (www.bgc-jena.mpg.de/geodb/projects), is also analysed. See Validation Report (VR) document for a more detailed information (SAF/LAND/UV/VR_MGPP/1.3).
The proposed algorithm to derive the MGPP product relies on Monteith's light use efficiency (LUE) concept (Monteith, 1972). This concept provides the theoretical basis for most models of GPP relying on optical remote sensing and considers daily GPP (g m-2 day-1) as proportional to the incoming photosynthetically active solar radiation (PAR) (MJ m-2 day-1), the fractional absorption of that flux (fAPAR) (dimensionless) and the light-use efficiency (ε) (g MJ-1) (Gilabert et al., 2015). The proposed algorithm first computes daily GPP (MDGPP) using the daily downward surface shortwave radiation flux (DIDSSF) (LSA‐203) and the daily fAPAR (MDFAPAR) (LSA‐425) (Martínez et al., 2018). The e is operationally parameterized as a maximum value (emax) depending on different biomes and is downregulated by an estimator related with water stress (Cws). The daily Cws is computed using the daily actual evapotranspiration LSA SAF product (DMET, LSA-302) and the reference ET, ET0, also provided by LSA SAF (DMETREF, LSA-303). Second, the MGPP product is calculated by averaging the 10 subsequent and consecutive daily observations (MDGPP). See Algorithm Theoretical Basis Document (ATBD) for a more detailed description (SAF/LAND/UV/ATBD_GPP/1.4 ).
The MGPP product is composed of 4 layers that depend on the MDGPP internal product: Layer 1 contains the 10-day GPP estimates. Layer 2 is considered a quality control information field and it refers to the number of days with fAPAR error higher than 0.18 over the 10-day period. Thus, this layer can be used to discard these areas. Layer 3 gives information about the associated errors, whereas layer 4 provides the number of MDGPP used in the MGPP computation. For layer 4, the greater the value the better the quality of the product. The MGPP is computed within the area covered by the MSG disk. The MGPP product is disseminated in HDF5 format.
Three MGPP products every calendar month are produced, with the first product running from the 1st till the 10th, the second one from the 11th till the 20th, and the last one running from the 21st till the last day of the month. Thus, there are 36 compositing MGPP products for each calendar year.
Automatic Quality Control (QC) is performed on MGPP product and the quality information is provided on a pixel basis (layer 2 and layer 4 of the MGPP product). QC contains general information about input data quality, and specific information related with the limits of application (number of cloud-free observations and unrealistic pixel values due to traces of snow, inland water or system failure). The error estimate defines the confidence level of the product on the basis of the theoretical model assumption.
Validation includes the assessment of the product in a systematic and statistically robust way representing global conditions. The validation results are highly significant and allow concluding that the MGPP product has reached a validation stage level 2 according to CEOS LPV criteria. Based on ground measurements, a compliance user requirement of 42% observations is obtained for the target user required accuracy and 67% of observations for the optimal requirements. An 80% of observations are below the threshold used requirement accuracy. Moreover, the MGPP product has shown a high confidence as compared to other RS GPP products.
Detailed validation studies are given in the Validation Report (SAF/LAND/UV/VR_MGPP/1.3) and in the Product User Manual (SAF/LAND/UV/PUM_MGPP/1.3).
Gilabert, M.A., et al., 2015. Daily GPP estimates in Mediterranean ecosystems by combining remote sensing and meteorological data. ISPRS Journal of Photogrammetry and Remote Sensing, 102, 184-197.
Martínez, B., Sanchez-Ruiz, S., Gilabert, M.A, Moreno, A., Campos-Taberner, M., García-Haro, F.J., Trigo, I.F., et al., 2018. Retrieval of daily gross primary production over Europe and Africa from an ensemble of SEVIRI/MSG products. International Journal of Applied Earth Observation and Geoinformation, 65, 124-136.
Monteith, J.L., 1972. Solar radiation and productivity in tropical ecosystems. Journal of Applied Ecology, 9, 747–766.