基于OCO-2遥感数据的新疆维吾尔自治区大气XCO2空间化研究

doi:10.16258/j.cnki.1674-5906.2024.02.007

生态环境学报 ›› 2024, Vol. 33 ›› Issue (2): 231-241.DOI: 10.16258/j.cnki.1674-5906.2024.02.007

• 研究论文【环境科学】 • 上一篇下一篇

基于OCO-2遥感数据的新疆维吾尔自治区大气XCO₂空间化研究

张杨¹(), 徐永明¹^,^*(), 卢响军², 莫亚萍¹, 吉蒙¹, 祝善友¹

1.南京信息工程大学遥感与测绘工程学院，江苏南京 210044
2.新疆生产建设兵团生态环境第一监测站，新疆乌鲁木齐 830011

收稿日期:2023-09-07 出版日期:2024-02-18 发布日期:2024-04-03
通讯作者: *徐永明。E-mail: xym30@263.net
作者简介:张杨（1999年生），女，硕士研究生，研究方向为大气环境遥感。E-mail: 20211211035@nuist.edu.cn
基金资助:
新疆生产建设兵团重点领域科技攻关计划(2022AB016);石河子市重点领域科技攻关计划(2022NY03)

Spatialization of Atmospheric XCO₂ in Xinjiang Uygur Autonomous Region based on OCO-2 Remote Sensing Data

ZHANG Yang¹(), XU Yongming¹^,^*(), LU Xiangjun², MO Yaping¹, JI Meng¹, ZHU Shanyou¹

1. School of Remote Sensing & Geomatics Engineering, Nanjing University of Information Science & Technology, Nanjing 210044, P. R. China
2. The First Ecological and Environment Monitoring Station of Xinjiang Production and Construction Corps, Urumchi 830011, P. R. China

Received:2023-09-07 Online:2024-02-18 Published:2024-04-03

摘要/Abstract

摘要：

二氧化碳（CO₂）是导致全球变暖最主要的温室气体，掌握准确的CO₂空间分布信息可以有效评估碳减排成效，对于推进碳达峰、碳中和工作具有重要意义。相比站点观测，碳卫星能够获取大尺度的CO₂分布信息，但是由于其幅宽较窄以及云覆盖的影响，大气CO₂卫星遥感数据存在大量缺值区域，不能获得空间连续的大气CO₂分布。以新疆维吾尔自治区为研究区，基于2019年OCO-2卫星大气二氧化碳柱平均干空气混合比（XCO₂）数据，结合气温、地形、植被、大气NO₂浓度等相关变量，综合对比了多元线性回归（MLR）、地理加权回归（GWR）、支持向量机（SVR）、随机森林（RF）、极端梯度提升树（XGBoost）和极端随机树（ERT）等方法在生成大气XCO₂空间连续数据中的表现。交叉验证结果表明，RF、XGBoost和ERT这3种集成学习模型精度明显优于SVR、GWR和MLR模型，其中ERT模型精度最高，决定系数R²为0.748，平均绝对误差σ_MAE为0.489×10⁻⁶。基于ERT模型生成了新疆2019年逐月大气XCO₂空间连续数据，该地区大气XCO₂空间差异和季节变化都很明显。在空间分布格局上总体呈现与地形“三山夹两盆”相似的空间分布特征，裸地XCO₂年均值最高，林地和草原的XCO₂年均值最低；季节变化特征与植被生长周期相似，春冬季浓度较高，夏秋季浓度较低，最高值出现在4月，最低值出现在8月，不同土地覆盖类型的大气XCO₂季节变化与全区域类似，其中草原和林地的季节差异最大，裸地季节差异最小。该研究为基于碳卫星遥感数据生成时空连续的大尺度XCO₂数据提供了参考和借鉴。

关键词: OCO-2, XCO₂, 卫星遥感, 集成学习模型, 空间化, 时空分布, 新疆维吾尔自治区

Abstract:

Carbon dioxide (CO₂) is a primary greenhouse gas that plays a crucial role in global warming. Accurate information on the spatial distribution of CO₂ is imperative for assessing the efficiency of carbon emission reduction, which is important for promoting carbon peaking and neutrality. Compared to station observations, carbon satellites can obtain large-scale information on CO₂ distribution. However, the narrow width and cloud coverage associated with atmospheric CO₂ satellite remote sensing data contribute to numerous gaps, hindering the acquisition of spatially continuous atmospheric CO₂ distribution. Using the Xinjiang Uygur Autonomous Region as the study area, this study synthetically compared the performance of several methods in generating spatially continuous atmospheric XCO₂data. Multiple linear regression (MLR) models, geographically weighted regression (GWR) models, support vector machines (SVR), random forests (RF), extreme gradient boosted trees (XGBoost), and extreme random trees (ERT) were employed to produce gap-free XCO₂ data based on the column-averaged dry-air mixing ratio of CO₂ (XCO₂) data derived from the 2019 OCO-2 satellite and auxiliary variables, such as temperature, topography, vegetation, and atmospheric NO₂ concentrations. The cross-validation results showed that the three integrated learning models, RF, XGBoost, and ERT, outperformed the SVR, GWR, and MLR models. The ERT model achieved the highest accuracy with an R² of 0.748 and a mean absolute error of 0.489×10⁻⁶. Based on the developed ERT model, spatially continuous atmospheric XCO₂ data was mapped monthly across Xinjiang in 2019. Significant spatial differences were observed in the atmospheric XCO₂ within the region. The spatial distribution pattern of XCO₂ exhibited a distinct correlation with the topographical characteristics commonly referred to as “three mountains and two basins.” Bare land exhibited the highest annual mean XCO₂, whereas woodland and grassland exhibited the lowest values. Seasonal fluctuations in XCO₂ coincided with the phenological cycles of vegetation, with peaks observed during spring and winter, and troughs during summer and autumn. The highest value was observed in April, whereas the lowest was observed in August. Across different land cover types, the seasonal variation in XCO₂ was similar to the regional patterns, with grassland and woodland experiencing the most pronounced seasonal disparities and bare land the least. This study provides a valuable framework for generating spatially and temporally continuous large-scale XCO₂ data using satellite XCO₂ products.

Key words: OCO-2, XCO₂, satellite remote sensing, integrated learning model, spatialization, spatiotemporal distribution, Xinjiang Uygur Autonomous Region

中图分类号:

张杨, 徐永明, 卢响军, 莫亚萍, 吉蒙, 祝善友. 基于OCO-2遥感数据的新疆维吾尔自治区大气XCO₂空间化研究[J]. 生态环境学报, 2024, 33(2): 231-241.

ZHANG Yang, XU Yongming, LU Xiangjun, MO Yaping, JI Meng, ZHU Shanyou. Spatialization of Atmospheric XCO₂ in Xinjiang Uygur Autonomous Region based on OCO-2 Remote Sensing Data[J]. Ecology and Environment, 2024, 33(2): 231-241.

图/表 12

图1 研究区概况文本所有地图基于自然资源部标准地图服务网站GS (2019) 1822号标准地图制作，底图边界无修改

Figure 1 Overview of the studied area

表1 数据信息

Table 1 Data information table

数据名称	数据来源	时间分辨率	空间分辨率	数据时间
XCO₂	OCO-2	16 d	1.29 km× 2.25 km	2019
NDVI	MODIS/MOD13A3	月	1 km×1 km	2019
气象再分析数据	ERA5-Land	月	0.1°×0.1°	2019
大气NO₂	Sentinel-5P	d	1.1132 km× 1.1132 km	2019
夜间灯光数据	NPP/VIIRS	月	500 m× 500 m	2019
DEM	SRTM/SRTMGL1_v003	‒	30 m×30 m	2000
土地覆盖	MODIS/MCD12Q1	‒	500 m×500 m	2019

图2 新疆维吾尔自治区2019年每月OCO-2观测数据空间覆盖

Figure 2 Spatial coverage of monthly OCO-2 observations in Xinjiang Uygur Autonomous Region in 2019

表2 多重共线性回归分析结果

Table 2 Results of multiple covariance regression analysis

月份	VIF
月份	NDVI	气温	风速	风向	高程	大气NO₂	夜间灯光
1	1.18	4.94	1.82	1.05	5.13	1.15	1.03
2	1.94	3.80	1.62	1.08	3.27	1.83	1.05
3	1.33	3.30	1.35	1.43	3.09	1.57	1.02
4	1.60	3.09	1.56	1.51	3.94	2.04	1.00
5	1.74	4.20	1.39	1.29	4.39	2.13	1.02
6	2.52	6.97	1.48	1.28	7.76	3.04	1.05
7	1.71	5.49	1.12	1.02	8.39	3.23	1.02
8	1.79	8.07	1.48	1.22	9.07	1.78	1.02
9	1.49	6.22	1.11	1.30	5.75	1.49	1.00
10	1.17	3.75	1.15	1.26	4.01	1.54	1.01
11	1.10	2.09	1.86	1.17	3.24	1.28	1.03
12	1.47	5.36	1.43	1.03	4.81	1.26	1.07

图3 6种模型的交叉验证精度箱型图

Figure 3 Box plot of cross-validation accuracy for 7 models

图4 模型估算值与遥感观测值散点图

Figure 4 Scatter plot of model estimates and remote sensing observations

表3 数据集不同区间的误差统计结果

Table 3 Error statistics for different intervals of the dataset

范围 (×10⁻⁶)	数量	σ_MBE (×10⁻⁶)	σ_MAE (×10⁻⁶)
395-405	22691	0.359	0.544
405-415	313536	−0.024	0.468
415-425	748	−1.80	1.90

图5 ERT模型变量重要性

Figure 5 Importance of ERT model variables

图6 新疆维吾尔自治区2019年大气XCO2空间分布特征

Figure 6 Spatial distribution characteristics of atmospheric XCO2 of Xinjiang Uygur Autonomous Region in 2019

图7 新疆维吾尔自治区2019年土地覆盖图及XCO2统计结果

Figure 7 Landcover map and XCO2 statistics results of Xinjiang Uygur Autonomous Region in 2019

图8 新疆维吾尔自治区2019年每月XCO2空间分布

Figure 8 Spatial distribution of monthly XCO2 in Xinjiang Uygur Autonomous Region in 2019

图9 不同土地覆盖类型的XCO2变化情况

Figure 9 XCO2 changes in different land cover types

参考文献 39

[1]	AUMANN H H, CHAHINE M T, GAUTIER C, et al., 2003. AIRS/AMSU/ HSB on the aqua mission: Design, science objectives, data products, and processing systems[J]. IEEE Transactions on Geoscience and Remote Sensing, 41(2): 253-264. DOI URL
[2]	BHATTACHARJEE S, CHEN J, 2020. Prediction of satellite-based column CO₂ concentration by combining emission inventory and LULC information[J]. IEEE Transactions on Geoscience and Remote Sensing, 58(12): 8285-8300. DOI URL
[3]	BOVENSMANN H, BURROWS J P, BUCHWITZ M, et al., 1999. SCIAMACHY: Mission objectives and measurement modes[J]. Journal of the Atmospheric Sciences, 56(2): 127-150. DOI URL
[4]	BREIMAN L, 2001. Random forests[J]. Machine Learning, 45(1): 5-32. DOI URL
[5]	BRUNSDON C, FOTHERINGHAM S, CHARLTON M, 1998. Geographically weighted regression[J]. Journal of the Royal Statistical Society: Series D (The Statistician), 47(3): 431-443.
[6]	CHEN H P, SUN J Q, 2019. Increased population exposure to extreme droughts in China due to 0.5 ℃ of additional warming[J]. Environmental Research Letters, 14(6): 064011. DOI
[7]	CHEN J, SHI X Y, GU L, et al., 2023. Impacts of climate warming on global floods and their implication to current flood defense standards[J]. Journal of Hydrology, 618(4): 129236. DOI URL
[8]	CHEN T Q, GUESTRIN C, 2016. Xgboost: A scalable tree boosting system[C]// Proceedings of the 22nd acm sigkdd international conference on knowledge discovery and data mining. San Francisco: Association for Computing Machinery: 785-794.
[9]	CRISP D, POLLOCK H R, ROSENBERG R, et al., 2017. The on-orbit performance of the Orbiting Carbon Observatory-2 (OCO-2) instrument and its radiometrically calibrated products[J]. Atmospheric Measurement Techniques, 10(1): 59-81. DOI URL
[10]	ELDERING A, WENNBERG P O, CRISP D, et al., 2017. The Orbiting Carbon Observatory-2 early science investigations of regional carbon dioxide fluxes[J]. Science, 358(6360): eaam5745. DOI URL
[11]	FALAHATKAR S, MOUSAVI S M, FARAJZADEH M, 2017. Spatial and temporal distribution of carbon dioxide gas using GOSAT data over IRAN[J]. Environmental Monitoring and Assessment, 189(12): 627. DOI PMID
[12]	GEURTS P, ERNST D, WEHENKEL L, 2006. Extremely randomized trees[J]. Machine Learning, 63(1): 3-42. DOI URL
[13]	GOLKAR F, AL-WARDY M, SAFFARI S F, et al., 2019. Using OCO-2 satellite data for investigating the variability of atmospheric CO₂ concentration in relationship with precipitation, relative humidity, and vegetation over Oman[J]. Water, 12(1): 101. DOI URL
[14]	GUO M, XU J W, WANG X F, et al., 2015. Estimating CO₂ concentration during the growing season from MODIS and GOSAT in East Asia[J]. International Journal of Remote Sensing, 36(17): 4363-4383. DOI URL
[15]	HOEGH-GULDBERG O, JACOB D, TAYLOR M, et al., 2019. The human imperative of stabilizing global climate change at 1.5 ℃[J]. Science, 365(6459): eaaw6974. DOI URL
[16]	HOFMANN D J, BUTLER J H, TANS P P, 2009. A new look at atmospheric carbon dioxide[J]. Atmospheric Environment, 43(12): 2084-2086. DOI URL
[17]	JONES M W, PETERS G P, GASSER T, et al., 2023. National contributions to climate change due to historical emissions of carbon dioxide, methane, and nitrous oxide since 1850[J]. Scientific Data, 10(1): 155. DOI PMID
[18]	KIEL M, O’DELL C W, FISHER B, et al., 2019. How bias correction goes wrong: measurement of XCO₂ affected by erroneous surface pressure estimates[J]. Atmospheric Measurement Techniques, 12(4): 2241-2259. DOI URL
[19]	KIM J B, SO J M, BAE D H, 2020. Global warming impacts on severe drought characteristics in Asia monsoon region[J]. Water, 12(5): 1360. DOI URL
[20]	LI J, JIA K, WEI X Q, et al., 2022. High-spatiotemporal resolution mapping of spatiotemporally continuous atmospheric CO₂ concentrations over the global continent[J]. International Journal of Applied Earth Observation and Geoinformation, 108: 102743. DOI URL
[21]	LIU Q W, WU S M, LEI Y L, et al., 2021. Exploring spatial characteristics of city-level CO₂ emissions in China and their influencing factors from global and local perspectives[J]. Science of the Total Environment, 754: 142206. DOI URL
[22]	LIU Y, WANG X F, GUO M, et al., 2012. Mapping the FTS SWIR L2 product of XCO₂ and XCH₄ data from the GOSAT by the Kriging method: A case study in East Asia[J]. International Journal of Remote Sensing, 33(10): 3004-3025. DOI URL
[23]	PARK C, JEONG S, PARK H, et al., 2021. Evaluation of the potential use of satellite-derived XCO₂ in detecting CO₂ enhancement in megacities with limited ground observations: A case study in Seoul using orbiting carbon Observatory-2[J]. Asia-Pacific Journal of Atmospheric Sciences, 57(2): 289-299. DOI
[24]	RAN Y H, LI X, 2019. TanSat: A new star in global carbon monitoring from China[J]. Science Bulletin, 64(5): 284-285. DOI PMID
[25]	SIABI Z, FALAHATKAR S, ALAVI S J, 2019. Spatial distribution of XCO₂ using OCO-2 data in growing seasons[J]. Journal of Environmental Management, 244: 110-118. DOI URL
[26]	SMOLA A J, SCHÖLKOPF B, 2004. A tutorial on support vector regression[J]. Statistics and Computing, 14(3): 199-222. DOI URL
[27]	WANG L B, ROHLI R V, LIN Q G, et al., 2022. Impact of extreme heatwaves on population exposure in China due to additional warming[J]. Sustainability, 14(18): 11458. DOI URL
[28]	WATANABE H, HAYASHI K, SAEKI T, et al., 2015. Global mapping of greenhouse gases retrieved from GOSAT Level 2 products by using a kriging method[J]. International Journal of Remote Sensing, 36(6): 1509-1528. DOI URL
[29]	YANG D X, LIU Y, CAI Z N, et al., 2018. First global carbon dioxide maps produced from TanSat measurements[J]. Advances in Atmospheric Sciences, 35(6): 621-623. DOI
[30]	YOKOTA T, YOSHIDA Y, EGUCHI N, et al., 2009. Global concentrations of CO₂ and CH₄ retrieved from GOSAT: First preliminary results[J]. Sola, 5: 160-163. DOI URL
[31]	ZENG Z C, LEI L P, STRONG K, et al., 2017. Global land mapping of satellite-observed CO₂ total columns using spatio-temporal geostatistics[J]. International Journal of Digital Earth, 10(4): 426-456. DOI URL
[32]	ZHANG M Q, LIU G J, 2023. Mapping contiguous XCO₂ by machine learning and analyzing the spatio-temporal variation in China from 2003 to 2019[J]. Science of The Total Environment, 858(Part 2): 159588. DOI URL
[33]	陈慧, 缪启龙, 买买提艾力·买买提依明, 等, 2015. 塔克拉玛干沙漠腹地春、夏季CO₂通量特征[J]. 干旱区地理, 38(3): 487-493.
	CHEN H, MIAO Q L, ALI M, et al., 2015. CO₂ flux characteristics of Taklimakan Desert in spring and summer[J]. Arid Land Geography, 38(3): 487-493.
[34]	陈良富, 尚华哲, 范萌, 等, 2021. 高分五号卫星大气参数探测综述[J]. 遥感学报, 25(9): 1917-1931.
	CHEN L F, SHANG H Z, FAN M, et al., 2021. Mission overview of the GF-5 satellite for atmospheric parameter monitoring[J]. National Remote Sensing Bulletin, 25(9): 1917-1931. DOI URL
[35]	付宏臣, 孙艳玲, 陈莉, 等, 2020. 基于AOD数据与GWR模型的2016年新疆地区PM_2.5和PM₁₀时空分布特征[J]. 环境科学学报, 40(1): 27-35.
	FU H C, SUN Y L, CHEN L, et al., 2020. Temporal and spatial distribution characteristics of PM_2.5 and PM₁₀ in Xinjiang region in 2016 based on AOD data and GWR model[J]. Acta Scientiae Circumstantiae, 40(1): 27-35.
[36]	刘良云, 陈良富, 刘毅, 等, 2022. 全球碳盘点卫星遥感监测方法、进展与挑战[J]. 遥感学报, 26(2): 243-267.
	LIU L Y, CHEN L F, LIU Y, et al., 2022. Satellite remote sensing for global stocktaking: methods, progress and perspectives[J]. National Remote Sensing Bulletin, 26(2): 243-267. DOI URL
[37]	孟倩文, 尹球, 2016. 中国区域CO₂多年时空变化的卫星遥感分析[J]. 遥感技术与应用, 31(2): 203-213.
	MENG Q W, YIN Q, 2016. Remote sensing analysis of multi-years spatial and temporal variation of CO₂ in China[J]. Remote Sensing Technology and Application, 31(2): 203-213.
[38]	莫露, 巫兆聪, 张熠, 2021. 中国XCO₂时空分布与影响因素分析[J]. 中国环境科学, 41(6): 2562-2570.
	MO L, WU Z C, ZHANG Y, 2021. Spatial and temporal variations of XCO₂ in China and its influencing factors analysis[J]. China Environmental Science, 41(6): 2562-2570.
[39]	肖钟湧, 陈颖锋, 林晓凤, 等, 2022. 基于多源卫星遥感数据的中国2003年-2018年CO₂时空变化研究[J]. 遥感学报, 26(12): 2486-2496.
	XIAO Z Y, CHEN Y F, LIN X F et al., 2022. The temporal and spatial variation of CO₂ column concentration over China from 2003 to 2018 based on multi-source satellite remote sensing data[J]. National Remote Sensing Bulletin, 26(12): 2486-2496. DOI URL

基于OCO-2遥感数据的新疆维吾尔自治区大气XCO₂空间化研究

Spatialization of Atmospheric XCO₂ in Xinjiang Uygur Autonomous Region based on OCO-2 Remote Sensing Data

RichHTML

PDF

可视化

摘要/Abstract

引用本文

使用本文

图/表 12

参考文献 39

相关文章 10

编辑推荐

Metrics

本文评价

[1]	袁茜, 傅开道, 陶雨晨, 张年, 杨丽莎. 澜沧江（云南段）水-气界面氧化亚氮释放通量时空分布特征及其影响因素研究[J]. 生态环境学报, 2024, 33(1): 54-61.
[2]	吴雅睿, 王美景, 王涛, 杨梅焕. 新冠疫情下NO₂时空变化特征——以陕西省为例[J]. 生态环境学报, 2023, 32(3): 514-524.
[3]	李姝亭, 胡冠九, 罗小三. 大气环境中全（多）氟烷基化合物（PFASs）的来源、分布及健康风险研究进展[J]. 生态环境学报, 2023, 32(12): 2103-2114.
[4]	刘明宇, 郑旭, 强丽媛, 李鲁华, 张若宇, 王家平. 1994－2020年中国农用薄膜使用量变化与农膜微塑料污染现状分析[J]. 生态环境学报, 2023, 32(11): 2050-2061.
[5]	孙正, 曹亚非, 王德彩, 刘峰, 宋效东, 张甘霖, 吴华勇. 近30年京津冀电镀场地时空演变特征及趋势预测[J]. 生态环境学报, 2023, 32(1): 183-194.
[6]	魏小锋, 韩红, 闫学军, 王在峰, 李圣增, 田勇, 梁第, 马明亮, 张桂芹. 基于卫星遥感与CMB模型的济南市冬季重污染过程PM_2.5溯源分析[J]. 生态环境学报, 2022, 31(6): 1175-1183.
[7]	贺斌, 胡茂川. 广东省各区县农业面源污染负荷估算及特征分析[J]. 生态环境学报, 2022, 31(4): 771-776.
[8]	王金杰, 赵安周, 胡小枫. 京津冀植被净初级生产力时空分布及自然驱动因子分析[J]. 生态环境学报, 2021, 30(6): 1158-1167.
[9]	郑诗禹, 张绿水, 郭晓敏, 黄子峻, 肖以华. 不同森林郁闭度环境内空气负氧离子的时空变化及环境影响要素研究——以广州帽峰山为例[J]. 生态环境学报, 2021, 30(11): 2204-2212.
[10]	王薇, 程歆玥, 胡春, 夏斯涵, 王甜. 城市街道峡谷PM_2.5时空分布特征与空气质量评价——以合肥市长淮街道为例[J]. 生态环境学报, 2021, 30(11): 2157-2164.