Abstract. Accurately quantifying the spatiotemporal distribution of the biological
component of CO2 surface–atmosphere exchange is necessary to improve
top-down constraints on China's anthropogenic CO2 emissions. We
provide hourly fluxes of CO2 as net ecosystem exchange (NEE;
µmol CO2 m−2 s−1) on a 0.25∘×0.25∘ grid by adapting the Vegetation, Photosynthesis, and
Respiration Model (VPRM) to the eastern half of China for the time period
from 2005 to 2009; the minimal empirical parameterization of the VPRM-CHINA
makes it well suited for inverse modeling approaches. This study diverges
from previous VPRM applications in that it is applied at a large scale to
China's ecosystems for the first time, incorporating a novel processing
framework not previously applied to existing VPRM versions. In addition, the
VPRM-CHINA model prescribes methods for addressing dual-cropping regions that
have two separate growing-season modes applied to the same model grid cell. We
evaluate the VPRM-CHINA performance during the growing season and compare to
other biospheric models. We calibrate the VPRM-CHINA with ChinaFlux and
FluxNet data and scale up regionally using Weather Research and Forecasting
(WRF) Model v3.6.1 meteorology and MODIS surface reflectances. When combined
with an anthropogenic emissions model in a Lagrangian particle transport
framework, we compare the ability of VPRM-CHINA relative to an ensemble mean
of global hourly flux models (NASA CMS – Carbon Monitoring System) to reproduce observations made at a
site in northern China. The measurements are heavily influenced by the
northern China administrative region. Modeled hourly time series using
vegetation fluxes prescribed by VPRM-CHINA exhibit low bias relative to
measurements during the May–September growing season. Compared to NASA CMS
subset over the study region, VPRM-CHINA agrees significantly better with
measurements. NASA CMS consistently underestimates regional uptake in the
growing season. We find that during the peak growing season, when the heavily
cropped North China Plain significantly influences measurements, VPRM-CHINA
models a CO2 uptake signal comparable in magnitude to the modeled
anthropogenic signal. In addition to demonstrating efficacy as a low-bias
prior for top-down CO2 inventory optimization studies using
ground-based measurements, high spatiotemporal resolution models such as the
VPRM are critical for interpreting retrievals from global CO2 remote-sensing platforms such as OCO-2 and OCO-3 (planned). Depending on the
satellite time of day and season of crossover, efforts to interpret the
relative contribution of the vegetation and anthropogenic components to the
measured signal are critical in key emitting regions such as northern China
– where the magnitude of the vegetation CO2 signal is shown to be
equivalent to the anthropogenic signal.
A modified Vegetation Photosynthesis and Respiration Model (VPRM) for the eastern USA and Canada, evaluated with comparison to atmospheric observations and other biospheric models
