Arid and semi-arid ecosystems exhibit a spatially complex biogeophysical structure. According to arid western special climate-vegetation characters, the fractional cover of photosynthetic vegetation （PV）, non-photos...Arid and semi-arid ecosystems exhibit a spatially complex biogeophysical structure. According to arid western special climate-vegetation characters, the fractional cover of photosynthetic vegetation （PV）, non-photosynthetic vegetation （NPV）, bare soil and water are unmixed, using the remote sensing spectral mixture analysis. We try the method to unmix the canopy funation structure of arid land cover in order to avoid the differentiation of regional vegetation system and the disturbance of environmental background. We developed a modified production efficiency model NPP-PEM appropriate for the arid area at regional scale based on the concept of radiation use efficiency. This model refer to the GLO-PEM and CASA model was driven with remotely sensed observations, and calculates not just the conversion efficiency of absorbed photosynthetically active radiation but also the carbon fluxes that determine net primary productivity （NPP）. We apply and validate the model in the Kaxger and Yarkant river basins in arid western China. The NPP of the study area in 1992 and 1998 was estimated based on the NPP-PEM model. The results show that the improved PEM model, considering the photosynthetical activation of heterogeneous functional vegetation, is in good agreement with field measurements and the existing literature. An accurate agreement （R2= 0.85, P〈0.001） between the estimates and the ground-based measurement was obtained. The spatial distribution of mountain-oasis-desert ecosystem shows an obvious heterogeneous carbon uptake. The results are applicable to arid ecosystem studies ranging from characterizing carbon cycle, carbon flux over arid areas to monitoring change in mountain-oasis-desert productivity, stress and management.展开更多
Spatially explicit modeling plays a vital role in land use/cover change and urbanization research as well as resources management;however,current models lack proper validation and fail to incorporate uncertainty into ...Spatially explicit modeling plays a vital role in land use/cover change and urbanization research as well as resources management;however,current models lack proper validation and fail to incorporate uncertainty into the formulation of model predictions.Consequently,policy makers and the general public may develop opinions based on potentially misleading research,which fails to allow for truly informed decisions.Here we use an uncertainty strategy of spatially explicit modeling combined with the series statistic of Kappa index for location and quantity to estimate the uncertainty of future predications and to determine model accuracy.We take the Beijing metropolitan area as an example to demonstrate the uncertainty in extrapolations of predictive land use change and urban sprawl with spatially explicit modeling at multiple resolutions.The sensitivity of scale effects is also discussed.The results show that an improvement in specification of location is more helpful in increasing accuracy as compared to an improvement in the specification of quantity at fine spatial resolutions.However,the spatial scale has great effects on modeling accuracy and correct due to chance tends to increase as resolution becomes coarser.The results allow us to understand the uncertainty when using spatially explicit models for land-use change or urbanization estimates.展开更多
基金National Project for Basic Research, No.2002CB412507 National key project of fundamental research, No.G1999043500 National Natural Science Foundation of China, No.90202002
文摘Arid and semi-arid ecosystems exhibit a spatially complex biogeophysical structure. According to arid western special climate-vegetation characters, the fractional cover of photosynthetic vegetation （PV）, non-photosynthetic vegetation （NPV）, bare soil and water are unmixed, using the remote sensing spectral mixture analysis. We try the method to unmix the canopy funation structure of arid land cover in order to avoid the differentiation of regional vegetation system and the disturbance of environmental background. We developed a modified production efficiency model NPP-PEM appropriate for the arid area at regional scale based on the concept of radiation use efficiency. This model refer to the GLO-PEM and CASA model was driven with remotely sensed observations, and calculates not just the conversion efficiency of absorbed photosynthetically active radiation but also the carbon fluxes that determine net primary productivity （NPP）. We apply and validate the model in the Kaxger and Yarkant river basins in arid western China. The NPP of the study area in 1992 and 1998 was estimated based on the NPP-PEM model. The results show that the improved PEM model, considering the photosynthetical activation of heterogeneous functional vegetation, is in good agreement with field measurements and the existing literature. An accurate agreement （R2= 0.85, P〈0.001） between the estimates and the ground-based measurement was obtained. The spatial distribution of mountain-oasis-desert ecosystem shows an obvious heterogeneous carbon uptake. The results are applicable to arid ecosystem studies ranging from characterizing carbon cycle, carbon flux over arid areas to monitoring change in mountain-oasis-desert productivity, stress and management.
基金supported by China Postdoctoral Science Foundation (Grant No.20070420630)National Basic Research Program of China (Grant Nos.2002CB412507,G19990435)
文摘Spatially explicit modeling plays a vital role in land use/cover change and urbanization research as well as resources management;however,current models lack proper validation and fail to incorporate uncertainty into the formulation of model predictions.Consequently,policy makers and the general public may develop opinions based on potentially misleading research,which fails to allow for truly informed decisions.Here we use an uncertainty strategy of spatially explicit modeling combined with the series statistic of Kappa index for location and quantity to estimate the uncertainty of future predications and to determine model accuracy.We take the Beijing metropolitan area as an example to demonstrate the uncertainty in extrapolations of predictive land use change and urban sprawl with spatially explicit modeling at multiple resolutions.The sensitivity of scale effects is also discussed.The results show that an improvement in specification of location is more helpful in increasing accuracy as compared to an improvement in the specification of quantity at fine spatial resolutions.However,the spatial scale has great effects on modeling accuracy and correct due to chance tends to increase as resolution becomes coarser.The results allow us to understand the uncertainty when using spatially explicit models for land-use change or urbanization estimates.
