Search

Terrestrial carbon dioxide (CO2) flux estimates in China using atmospheric inversion method are beset with considerable uncertainties because very few atmospheric CO2 concentration measurements are available. In order to improve these estimates, nested atmospheric CO2 inversion during 2002–2008 is performed in this study using passenger aircraft-based CO2 measurements over Eurasia from the Comprehensive Observation Network for Trace gases by Airliner (CONTRAIL) project. The inversion system includes 43 regions with a focus on China, and is based on the Bayesian synthesis approach and the TM5 transport model. The terrestrial ecosystem carbon flux modeled by the Boreal Ecosystems Productivity Simulator (BEPS) model and the ocean exchange simulated by the OPA-PISCES-T model are considered as the prior fluxes. The impacts of CONTRAIL CO2 data on inverted China terrestrial carbon fluxes are quantified, the improvement of the inverted fluxes after adding CONTRAIL CO2 data are rationed against climate factors and evaluated by comparing the simulated atmospheric CO2 concentrations with three independent surface CO2 measurements in China. Results show that with the addition of CONTRAIL CO2 data, the inverted carbon sink in China increases while those in South and Southeast Asia decrease. Meanwhile, the posterior uncertainties over these regions are all reduced (2–12%). CONTRAIL CO2 data also have a large effect on the inter-annual variation of carbon sinks in China, leading to a better correlation between the carbon sink and the annual mean climate factors. Evaluations against the CO2 measurements at three sites in China also show that the CONTRAIL CO2 measurements may have improved the inversion results.

In this study, we establish a nested atmospheric inversion system with a focus on China using the Bayesian method. The global surface is separated into 43 regions based on the 22 TransCom large regions, with 13 small regions in China. Monthly CO2 concentrations from 130 GlobalView sites and 3 additional China sites are used in this system. The core component of this system is an atmospheric transport matrix, which is created using the TM5 model with a horizontal resolution of 3° × 2°. The net carbon fluxes over the 43 global land and ocean regions are inverted for the period from 2002 to 2008. The inverted global terrestrial carbon sinks mainly occur in boreal Asia, South and Southeast Asia, eastern America and southern South America. Most China areas appear to be carbon sinks, with strongest carbon sinks located in Northeast China. From 2002 to 2008, the global terrestrial carbon sink has an increasing trend, with the lowest carbon sink in 2002. The inter-annual variation (IAV) of the land sinks shows remarkable correlation with the El Niño Southern Oscillation (ENSO). The terrestrial carbon sinks in China also show an increasing trend. However, the IAV in China is not the same as that of the globe. There is relatively stronger land sink in 2002, lowest sink in 2006, and strongest sink in 2007 in China. This IAV could be reasonably explained with the IAVs of temperature and precipitation in China. The mean global and China terrestrial carbon sinks over the period 2002–2008 are −3.20 ± 0.63 and −0.28 ± 0.18 PgC yr−1, respectively. Considering the carbon emissions in the form of reactive biogenic volatile organic compounds (BVOCs) and from the import of wood and food, we further estimate that China's land sink is about −0.31 PgC yr−1.

To understand the causes of the past water cycle variations and the influence of climate variability on the streamflow, lake storage, and flood potential, we analyze the changes in streamflow and the underlying drivers in four typical watersheds (Gaosha, Meigang, Saitang, and Xiashan) within the Poyang Lake Basin, based on the meteorological observations at 79 weather stations, and datasets of streamflow and river level at four hydrological stations for the period of 1961-2000. The contribution of different climate factors to the change in streamflow in each watershed is estimated quantitatively using the water balance equations. Results show that in each watershed, the annual streamflow exhibits an increasing trend from 1961–2000. The increases in streamflow by 4.80 m3 s−1 yr−1 and 1.29 m3 s−1 yr−1 at Meigang and Gaosha, respectively, are statistically significant at the 5% level. The increase in precipitation is the biggest contributor to the streamflow increment in Meigang (3.79 m3 s−1 yr−1), Gaosha (1.12 m3 s−1 yr−1), and Xiashan (1.34 m3 s−1 yr−1), while the decrease in evapotranspiration is the major factor controlling the streamflow increment in Saitang (0.19 m3 s−1 yr−1). In addition, radiation and wind contribute more than actual vapor pressure and mean temperature to the changes in evapotranspiration and streamflow for the four watersheds. For revealing the possible change of streamflow due to the future climate change, we also investigate the projected precipitation and evapotranspiration from of the Coupled Model Intercomparison Project phase 3 (CMIP3) under three greenhouse gases emission scenarios (SRESA1B, SRESA2 and SRESB1) for the period of 2061–2100. When the future changes in the soil water storage changes are assumed ignorable, the streamflow shows an uptrend with the projected increases in both precipitation and evapotranspiration (except for the SRESB1 scenario in Xiashan watershed) relative to the observed mean during 1961–2000. Furthermore, the largest increase in the streamflow is found at Meigang (+4.31%) and Xiashan (+3.84%) under the SRESA1B scenario, while the increases will occur at Saitang (+6.87%) and Gaosha (+5.15%) under the SRESB1 scenario.

This study aimed to exploit the use of deep learning networks in the retrieval of the biophysical and biochemical parameters of vegetation canopies. Convolutional Neural Network (CNN), network with only fully connected layers, referred as dense network (DNN), and Autoencoder (AE) were investigated to retrieve leaf area index (LAI) and leaf chlorophyll content. Hyperspectral data simulated by the coupled PROSPECT and SAIL model were used for training and validation. The real CASI hyperspectral data in 50 spectral channels ranging from 522.4 nm to 894.2 nm collected over three agricultural crop fields during the growing season of 2001 were used, together with the in-situ measured LAI and leaf chlorophyll content, as independent test set. Occlusion analysis was also employed to determine the important spectral bands at which reflectance made more contributions to the retrieval with a CNN and interpret the latent variables of the AE. Satisfactory results from these deep learning networks were obtained, compared with ground truth. The DNN with the input of the vegetation indices sensitive to LAI and leaf chlorophyll content (MTVI2 and TCARI/OSAVI) generated the best results with R2 of 0.86 for LAI and 0.55 for leaf chlorophyll content.