Abstract: Abstract: To mitigate the global warming caused by the excessive emission of CO2, Carbon Capture and Storage (CCS) techniques have been proposed to reduce the concentrations of atmospheric CO2 and to slow down the change of climate. However, everything has two sides for there is a risk of the CO2 leakage from the stored sites that may impact the surrounding environment significantly. Therefore, the monitoring of CO2 leaking spots has become a crucial issue to be solved in applying CCS. It is therefore needed to develop a large-scale, quick and highly efficient method for detecting the CO2 leakage on the surface of the sequestrating fields. Considering that the hyperspectral remote sensing technique can monitor the slight changes of surface vegetation by spectral feature analysis, this paper is dedicated to studying the impacts of the slight CO2 leakage stress on the surface vegetation through simulating experiment in the field. The experiment was carried out from May 2008 to August 2008 at the Sutton Bonington Campus（52.8 N, 1.2W）of the University of Nottingham. Beans (Vicia faba cv Clipper) were sowed by hand on June 4th, 2008. From July 4th on, the controlled CO2 was injected into the soil at a rate of 1L/min, and the concentrations of CO2 in the soil were measured every day in the field. Additionally, the chlorophyll contents and spectral data of beans were measured one time every week in the laboratory. The results showed that when the concentrations of soil CO2 was under 15%, there was no significant difference for chlorophyll contents between the control and stressed beans (P>0.1). However, when the concentrations of soil CO2 was above 15%, there was a great significant difference for chlorophyll contents between control and stressed beans（P<0.001）. As the time passed by, the experimented beans became premature senescent, experienced fallen leaves, and even died. The spectral data were processed by the continuum removal method and the results indicated that in the green bands the spectral reflectance gradually increased as the CO2 concentrated in the soil; nevertheless, in other bands there were no apparent and stable rules that could be derived from the spectral analysis for the whole growth stage of beans. According to the spectral feature analysis of beans under the stress of CO2 leakage, a new index (Area(510-590nm)) was designed to identify the beans. The experiment results showed that the Area (510-590nm) index was able to identify the stressed beans when the CO2 concentrations in the soil were above 15%. However, the index was unable to identify the stressed beans when the CO2 concentration in the soil was fewer than 15%. It can be concluded that the research is of great importance and has application value for detecting the leakage spots, monitoring and evaluating land-surface ecology under CO2 leakage stress.