锡林郭勒草原是我国四大草原之一，也是内蒙古主要的天然草场。近年来，在各种因素影响下，锡林郭勒草原区生态系统发生重大变化，导致草原区碳储量也发生改变。本文尝试应用InVEST模型中的碳储量模块对2013年锡林郭勒草原碳储量进行研究，从而探究InVEST模型在碳储量研究中的适用性。结果表明：（1）锡林郭勒草原各草地类型的地下碳储量明显多于地上碳储量；（2）草甸草原碳储量>典型草原碳储量>荒漠草原碳储量；（3）应用InVEST模型计算不同草地类型地上和地下碳储量，草甸草原分别为196.7 g?m?2和1385.2 g?m?2；典型草原的分别为133.4 g?m?2和688.9 g?m?2；荒漠草原的分别为56.6 g?m?2和301 g?m?2。应用InVEST模型计算出来的碳储量值与前人采用实地采样得出的结果基本一致，所以InVEST模型对于草原区碳储量的研究具有一定的适用性。“,”Background, aim, and scope Xilin Gol Grassland, as one of the four prairies in China, is primary natural grassland in Inner Mongolia. Recently, ecosystem has been undergone great change in this area, on account of the effects of various factors, which lead to change of carbon storage in steppe accordingly. As large carbon storage, grassland plays an important role in terrestrial carbon cycling. Evaluating the carbon storage of grassland ecosystem accurately and its dynamic changes are conducive to predict global climate change, the relationship between grassland ecosystem and sustainable of grassland resources. Due to some factors, application of conventional method for determining the grassland carbon storage are not accurate. Thus InVEST model is used in thesis to attempt to quantitatively evaluate carbon storage of Xilin Gol Grassland in 2013, so as to explore the applicability of InVEST model to study carbon reserve in steppe.Materials and methods Compared toother models, using carbon storage module in InVEST model to calculate the grassland carbon stocks of Xilin Golonly needs to input the initial spatial data, and then run the module will be able to output a large number of target data. Running the InVEST model to estimate the carbon storage of grassland ecosystem, which needs two basic datas: the current land use and land cover (LULC) and the table of carbon pool. The LULC is raster dataset which interpreted by ArcGIS. The data of carbon pool can be found in some relevant research, also can be extracted in relevant habitats by experiment, if it still can’t get relevant data, according to other carbon pool or set the default to 0. The collected data type-in on the basis of carbon model, we can get the aboveground carbon storage, underground carbon storage, soil carbon storage and total carbon storage, on the basis of ifgure, we can read the related ifelds of carbon distribution in the study area.Results Result show that (1) Through the research to carbon reserve of study area, we can konw that, the carbon distribution is closely related to vegetation types, and the underground carbon signiifcantly more than the aboveground, aboveground and underground carbon storage of Meadow steppe are 196.7 g?m?2 and 1385.2 g?m?2; aboveground and underground carbon storage of Typical grassland are 133.4 g?m?2 and 688.9 g?m?2; aboveground and underground carbon storage of Desert grassland are 56.6 g?m?2 and 301 g?m?2. (2) As an important part of grassland vegetation carbon storage, most of the biomass carbon is distributed in underground. The underground carbon distribution and output ifgure of Xilin Gol Grassland show that the carbon content in study area with 0 — 1400 g?m?2, the distribution of underground biomass carbon is closely related to vegetation types, the carbon storage of the meadow steppe area is highest, the typical grassland area is moderate and desert grassland carbon content at least. (3) The aboveground biomass carbon plays a decisive role in ecosystem productivity. The distribution of carbon still exists great correlation with vegetation types. Meadow steppe carbon reserves is the largest output with 120 — 200 g?m?2; Typical steppe carbon relatively less than meadow on the aboveground, with 80 —120 g?m?2; Desert steppe carbon is minimum of the aboveground, under 80 g?m?2. (4) Soil organic carbon pool is the largest of the earth surface ecosystem, its carbon storage has great relationship with vegetation net primary productivity, if vegetation net primary productivity is high, so with carbon storage, the trend of soil carbon reserve is Meadow grassland>Typical steppe>Desert steppe. (5) Grassland total carbon including aboveground carbon, underground carbon and carbon in the soil, the carbon of underground and carbon in soil are big part of grassland carbon. The total carbon of Xilin Gol league with 20 — 2350 g?m?2, and the distribution of carbon is different in different vegetation type areas. The carbon of Meadow steppe with 850 — 2350 g?m?2, the carbon of Typical grassland ranks second, with 150 — 850 g?m?2, due to less vegetation distribution, so the carbon of Desert grassland is less than 150 g?m?2.Discussion Although it is timely and effective to calculate carbon reserve of different vegetation type using InVEST model in study area, but there are some problems, such as the validity and accuracy of the initial data, lack of basic data, and dififcult to collect, which affecting the accuracy of the results, thus to establish complete and effective spatial database is an urgent task.Conclusions The result of carbon storage calculated by InVEST model is accordance with ifeld sampling by predecessor, showing that InVEST model for the study of carbon storage in grassland area has some applicability. It can not only avoid the complexity and restriction of the ifeld survey, but also verify the applicability of the model.Recommendations and perspectives We must make full use of data preparation for researching carbon reserves by InVEST model, on the basis of study area to interpret the image, we can get the LUCC ifgure accurately. Making full use of InVEST model, can not only calculate the amount of carbon storage accurately, but also avoid the complexity of the ifeld sampling survey, it has a great signiifcance to formulate environmental policy.