Miia Eskelinen, Sari Metsämäki The Finnish Environment Institute Geoinformatics and Land use division P.O.Box 140, FI-00251 Helsinki, Finland miia.eskelinen@ymparisto.fi sari.metsamaki@ymparisto.fi ABSTRACT The objective of this work is to evaluate the use of the Medium Resolution Imaging Spectrometer (MERIS) data for seasonal snow cover monitoring specifically in the boreal forest belt. For this purpose, we tuned an existing method for fractional snow cover mapping in order to produce snow maps from MERIS imagery. The method was originally developed at the Finnish Environment Institute (SYKE), where it is successfully used to provide frequent Snow Covered Area (SCA) maps from Terra/MODIS data. The possibility to switch between MERIS and successive multi-spectral optical sensors could ensure the snow service continuity and sufficient data supply for snow monitoring and even enable improved accuracy in snow map production. We found that MERIS visible channels suit well for SCA- mapping. 1. INTRODUCTION Snow accumulation and melting is an essential part of the hydrologic cycle in the boreal zone. There is a need to monitor the melting process by regular mapping of snow-covered areas during the melting period. Earth observation provides a spatially and temporally effective means to obtain information on the snow cover extent in addition to the traditional in-situ weather station and snow-gauging network. In order to obtain the fraction of Snow Covered Area (SCA), a reflectance model-based method SCAmod [1] was developed and implemented into operational use at the Finnish Environment Institute (SYKE) in 2001. It was designed to best perform for boreal forest areas [2]. High accuracy is usually required for hydrological applications, particularly when models are applied at regional scales across medium and small size drainage basins. Typically, snow patchiness is included as a model parameter. Since 2003, data provided by SCAmod have been successfully assimilated to the Finnish nationwide operational hydrological modelling system improving the performance of run-off and river discharge forecasts provided by the models [3]. This kind of permanent activity requires sustainable availability of Earth observation data. This is a motivation for experiments on snow mapping capacity of different sensors. In SCAmod, the reflectance from a target area is expressed as a function of SCA. Average effective forest canopy transmissivity (a priori-information, generated from EO-data) for each unit area and generally applicable average reflectance values for wet snow, snow-free ground and dense forest canopy are applied as model parameters [1]. This approach enables the employment of the method for large boreal areas with a tolerable effort. Employment of transmissivity is beneficial as it allows SCA estimation in various kinds of areas, both forested and non-forested. The method is applicable to a variety of optical sensors; switching between sensors requires only tuning the values of the three contributing reflectances. Also the transmissivity is slightly sensor-dependent and should therefore be calculated for each sensor. In this study, we describe the principles of the SCAmod modification for MERIS-data. This work requires acquisition of spectra for relevant reflectance contributors. This is the primary function of Analytical Spectral Devices (ASD) Field Spec Pro JR spectroradiometer measurements described in this paper. The success of the method modification is demonstrated with MERIS-derived snow cover maps. 2. STUDY AREA AND DATA SETS Northern Finland represents both boreal forests and tundra. The landscape is relatively flat and the forest evolves from a consistent closed canopy in the south, to a patchy mosaic of open canopy forest approaching the northern tree line. The landscape consists of multilayer vegetation covered with a thick seasonal snow pack. In Figure 1, a typical coniferous forest located in Northern Finland is depicted. Finland has a comprehensive in situ measurement network for snow parameters (160 snow courses and 700 weather stations). The boreal zone is characterized by seasonal snow cover, which has a significant influence on the interactive Earth's surface and atmosphere system. For this reason, seasonal snow cover is a sensitive climate change indicator also in regional scale. In Finland, snow pack is also an important temporary fresh water THE USE OF MERIS SPECTROMETER DATA IN SEASONAL SNOW MAPPING _____________________________________________________ Proc. ‘Envisat Symposium 2007’, Montreux, Switzerland 23–27 April 2007 (ESA SP-636, July 2007)