『Abstract
　An important function of riparian ecosystems, usually characterized as nitrate-loaded wetland environments, is nitrogen removal by denitrification. Riparian buffer zones around large dams and water reservoirs are also recognized as hotspots for emission of nitrogen (N2) and nitrous oxide (N2O), the latter being a strong greenhouse gas. Research has proven that land use has an important effect on soil denitrification. A spatial landscape-scale approach for analyzing denitrification processes and land use effects can therefore be considered important for an adequate assessment and management of NO3^- losses and N2O emissions in riparian ecosystems.
　In this study, we couple a soil denitrification process model with remote sensing data and techniques to analyze the spatial and temporal dynamics of soil denitrification in the riparian area of the Guanting reservoir, an important water supply of Beijing, China. SPOT-5 and Landsat TM5 satellite data were used to interpret the spatial land surface information and derive model parameters. A laboratory-scale anaerobic incubation experiments was used to estimate the soil denitrification model parameters for the different soil types. Modeling results were compared and validated with data from a nearby experimental N2O emission research site. The overall average soil denitrification rate (SDR) of the Guanting riparian basin was 32.45 mg N m^-2 d^-1 during the simulation period from March to September 2007, with a maximum of 370.49 mg N m^-2 d^-1 appeared in August and the minimum of 0.02 mg N m^-2 d^-1 in March. Bottomland and wetlands had large SDR's, with an average daily rate of 80.20 and 136 mg N m^-2 d^-1 respectively. Forest, grassland and shrub showed lower values, with average daily rates of 25.21, 18.77 and 16.59mg N m^-2 d^-1 respectively. The modeling results also indicated that farmland and orchards had a relative high SDR (34.09 and 33.25 mg N m^-2 d^-1 respectively), with large fluctuations observed between June and August due to agricultural practices. As soil denitrification rates and N2 and N2O emissions showed to be strongly correlated to the different land use practices, this could be taken into consideration when planning best management strategies for non-point source pollution control and greenhouse gas mitigation.

Keywords: Denitrification; Riparian zone; Land use; Remote sensing; Reservoir catchment』

1. Introduction
2. Study area and database
　2.1. Study area characteristics
　2.2. Datasets
　　2.2.1. Meteorological data
　　2.2.2. Soil survey data
　　2.2.3. Remote sensing data
3. Methods
　3.1. Model description
　　3.1.1. Soil denitrification model
　　3.1.2. Soil moisture and heat sub-model
　　　3.1.2.1. Soil water sub-model
　　　3.1.2.2. Soil temperature sub-model
　3.2. Spatial parameters inversion
　3.3. Laboratory denitrification experiment
　3.4. Field riparian experiment
　3.5. Parameter optimization
4. Results and discussion
　4.1. Parameters calibration
　4.2. Model validation at the riparian buffer zone basin scale
　4.3. Spatially explicit modeling of soil denitrification rates
　4.4. Land use effects on soil denitrification
　　4.4.1. Spatial structure analysis of Guanting riparian basin
　　4.4.2. Land use effects on Nitrogen losses by denitrification
5. Conclusions
Acknowledgment
References