Approximation method for time-integral of photosynthesis for NPP estimation using remote sensing data: Case study in Mongolia
2005
Xiong, Y.(Nara Women's Univ. (Japan)) | Muramatsu, K. | Hirata, M. | Oishi, K. | Kaihotsu, I. | Takamura, T. | Furumi, S. | Fujiwara, N.
A net primary production (NPP) estimation algorithm based on data from temperate vegetation in Japan was developed by Furumi et al. It utilizes the change in the vegetation index (VIPD), and photosynthetically active radiation (PAR) over time to estimate NPP with the non-linear relationship between the NPP and the PAR. To employ this algorithm to estimate NPP from satellite data on a global scale requires time series PAR data, but such data are unavailable on a global scale. Data of daily mean global solar radiation (SR) exposure over 24 hours are available, and daily mean PAR over 24 hours can be calculated from daily mean global SR exposure over 24 hours. However, when daily mean PAR over 24 hours was used, NPP was overestimated. The relationship between PAR and photosynthesis is not linear, and plants photosynthesize only during the daytime. We therefore propose the daily or monthly mean PAR over the effective daylength for vegetation photosynthesis, which is calculated from daily or monthly mean PAR. The effective daylength for vegetation photosynthesis is defined as the time between sunrise and sunset minus 2.0 hours. The NPP estimated from the daily or monthly mean PAR over the effective daylength for vegetation photosynthesis changes by +-% when the effective daylength for vegetation photosynthesis changes +- 25% of 2.0 hours from 1.5 hours to 2.5 hours. We applied the monthly mean PAR over the effective daylength for vegetation photosynthesis to estimate the NPP from Landsat ETM + data for a semi-arid area of Mongolia. The estimated NPP had an estimation error of 26%. The results were compared with ground measurement data, which have a 20% measurement error. The values agreed within the ranges of error. In the study area, average estimated NPP was 0.056+-0.015 [kgCO2/(m(2).month)].
Показать больше [+] Меньше [-]