Mapping Regional Forest Evapotranspiration and Photosynthesis by Coupling Satellite Data with Ecosystem Simulation

Abstract

Annual evapotranspiration (ET) and net photosynthesis (PSN) were estimated for a mountainous 28 x 55 km region of predominantly coniferous forests in western Montana. A simple geographic information system integrated topographic, soils, vegetation, and climate data at a 1.1—km scale defined by the satellite sensor pixel size. Leaf area index (LAI) of the forest was estimated with data from the NOAA (National Oceanic and Atmospheric Administration) Advanced Very High Resolution Radiometer (AVHRR). Daily microclimate of each cell was estimated from ground and satellite data and interpolated using MT—CLIM, a mountain microclimate simulator. A forest ecosystem simulation model, Forest—BGC, was used to calculate ET and PSN daily for each cell. Ranges of estimated LAI (4—15), ET (25—60 cm/yr), and PSN (9—20 Mg°ha—1°yr—1) across the landscape follow the trends expected in both magnitude and spatial pattern. These estimates compared well with field measurements of related variables, although absolute validation of these predictions is not possible at large spatial scales.

Comparison of methods toward multi-scale forest carbon mapping and spatial uncertainty analysis: combining national forest inventory plot data and landsat TM images

“…Multiple approaches may be used to estimate forest carbon including (1) direct conversion of forest inventory plot data directly into forest biomass estimates and then carbon estimates (Fang et al 2001;Nowak and Crane 2002;Peltoniemi et al 2006); (2) use of forest growth models (Ryan 1991;Neilson et al 2007); (3) use of process models (Chen et al 2003;Running et al 1989); and (4) combining remotely sensed image data with forest inventory plot data (Tan et al 2007;Wang et al 2009;Zheng et al 2007). These methods all have their advantages and disadvantages.…”

“…These methods all have their advantages and disadvantages. Converting forest inventory plot data into forest carbon provides excellent local spatial resolution (Fang et al 2001); forest growth models provide excellent temporal resolution (Neilson et al 2007); and process models are based on biophysical parameters and account for forest carbon accumulation (Running et al 1989). However, only the use of remotely sensed images in tandem with forest inventory plot data provides spatially and temporally explicit estimates of forest carbon stocks (Wang et al 2009).…”

“…Remotely sensed image data provide reflectance information that is often correlated with forest attributes such as tree diameter and height (Chen et al 2003;Dong et al 2003;Myneni et al 2001;Running et al 1989;Tan et al 2007;Zheng et al 2007). The combination of these two sources of data provides the potential to increase the precision and resolution of forest carbon estimates.…”

Regional hydrologic and carbon balance responses of forests resulting from potential climate change

“…Current ET ranges from 25 to 55 cm; but projections show much of the region will have ET of 70 cm. Running et al (1989) simulated annual ET to within 10% of ET estimated by an annual precipitation minus streamflow hydrologic balance of two 180 km 2 gauged watersheds in this study area. This increase in ET will obviously have a feedback effect on atmospheric moisture and possibly cloud development.…”

Identifying structural self-similarity in mountainous landscapes

“…This work is part of an ongoing project on the ecological process modeling of forest evapotranspiration and primary productivity by investigators at NASA-Ames Research Center and the Universities of Montana and Toronto (Band et al, 1991;Running et al, 1989).…”