Ecoclimap, a new complete surface parameter global dataset at a 1-km resolution, is presented. It is intended to be used to initialize the soil-vegetation-atmosphere transfer schemes (SVATs) in meteorological and climate models (at all horizontal scales). The database supports the ''tile'' approach, which is utilized by an increasing number of SVATs. Two hundred and fifteen ecosystems representing areas of homogeneous vegetation are derived by combining existing land cover maps and climate maps, in addition to using Advanced Very High Resolution Radiometer (AVHRR) satellite data. Then, all surface parameters are derived for each of these ecosystems using lookup tables with the annual cycle of the leaf area index (LAI) being constrained by the AVHRR information. The resulting LAI is validated against a large amount of in situ ground observations, and it is also compared to LAI derived from the International Satellite Land Surface Climatology Project (ISLSCP-2) database and the Polarization and Directionality of the Earth's Reflectance (POLDER) satellite. The comparison shows that this new LAI both reproduces values coherent at large scales with other datasets, and includes the high spatial variations owing to the input land cover data at a 1-km resolution. In terms of climate modeling studies, the use of this new database is shown to improve the surface climatology of the ARPEGE climate model.

[1] This study investigates the performances of four major global Leaf Area Index (LAI) products at 1/11.2°spatial sampling and a monthly time step: ECOCLIMAP climatology, GLOBCARBON (from SPOT/VEGETATION and ATSR/AATSR), CYCLOPES (from SPOT/VEGETATION) and MODIS Collection 4 (main algorithm, from MODIS/TERRA). These products were intercompared during the 2001-2003 period over the BELMANIP network of sites. Their uncertainty was assessed by comparison with 56 LAI reference maps derived from ground measurements. CYCLOPES and MODIS depict realistic spatial variations at continental scale, while ECOCLIMAP poorly captures surface spatial heterogeneity, and GLOBCARBON tends to display erratic variations. ECOCLIMAP and GLOBCARBON show the highest frequency of successful retrievals while MODIS and CYCLOPES retrievals are frequently missing in winter over northern latitudes and over the equatorial belt. CYCLOPES and MODIS describe consistent temporal profiles over most vegetation types, while ECOCLIMAP does not show any interannual variations, and GLOBCARBON can exhibit temporal instability during the growing season over forests. The CYCLOPES, MODIS, and GLOBCARBON LAI values agree better over croplands and grasslands than over forests, where differences in vegetation structure representation between algorithms and surface reflectance uncertainties lead to substantial discrepancies between products. CYCLOPES does not reach high enough LAI values to properly characterize forests. In contrast, the other products have sufficient dynamic range of LAI to describe the global variability of LAI. Overall, CYCLOPES is the most similar product to the LAI reference maps. However, more accurate ground measurements and better representation of the global and seasonal variability of vegetation are required to refine this result.

International audiencea r t i c l e i n f o Essential climate variables such as LAI or FAPAR are required for the monitoring, understanding and modeling of land surfaces at the global scale. While several products were already developed from the current medium resolution sensors, the few validation exercises currently achieved highlighted significant discrepancies and inconsistencies. The objective of this study is to develop improved global estimates of LAI, FAPAR and FCOVER variables by capitalizing on the development and validation of already existing products. In a first step, the performances of the MODIS, CYCLOPES, GLOBCARBON and JRC-FAPAR products were reviewed. The MODIS and CYCLOPES products were then selected since they provide higher level of consistency. These products were fused to generate the improved LAI, FAPAR and FCOVER values that were later scaled to closely match their expected range of variation. Finally, neural networks were trained to estimate these fused and scaled products from SPOT-VEGETATION top of canopy directionally normalized reflectance values. The resulting GEOV1 products are associated to quality control flags as well as quantitative estimates of uncertainties. Performances of the GEOV1 products are finally evaluated in a companion paper. The GEOV1 products are freely available to the community at www.geoland2.eu from 1999 up to present, globally at 1/112° spatial sampling grid at the dekadal time step

No abstract