Please wait a minute...
img

官方微信

遥感技术与应用  2021, Vol. 36 Issue (1): 55-64    DOI: 10.11873/j.issn.1004-0323.2021.1.0055
一带一路遥感监测专栏     
基于GEE平台的海岛地表覆盖提取及变化监测—以苏拉威西岛为例
付甜梦1,2(),张丽1,3(),陈博伟1,闫敏1
1.中国科学院空天信息创新研究院 数字地球重点实验室,北京 100094
2.中国科学院大学,北京 100049
3.海南省地球观测重点实验室,海南 三亚 572029
Monitoring of Land Cover Change based on Google Earth Engine Platform: A Case Study of Sulawesi Island
Tianmeng Fu1,2(),Li Zhang1,3(),Bowei Chen1,Min Yan1
1.Key Laboratory of Digital Earth Science,Aerospace Information Research Institute,Chinese Academy of Sciences,Beijing 100094,China
2.University of Chinese Academy of Sciences,Beijing 100049,China
3.Key Laboratory of Earth Observation of Hainan Province,Sanya 572029,China
 全文: PDF(5505 KB)   HTML
摘要:

随着海洋在国家政治、经济、资源等方面重要性的提升,对海岛开发利用、管理和保护等具有重要意义的海岛地表覆盖研究逐渐受到关注。针对传统的大范围地表覆盖提取速度慢、效率低的问题,基于Google Earth Engine(GEE)地学大数据平台,以作为生态重点保护区的印度尼西亚的苏拉威西岛为研究区,开展海岛地表覆盖提取及变化监测研究。基于GEE平台对Landsat TM/OLI遥感影像进行辐射校正、影像去云、镶嵌裁剪等预处理,构建海岛地表覆盖分类体系(包括人造地表、裸地、水体、湿地、草地、林地、耕地7个类别),采用随机森林(Random Forest, RF)算法提取2000~2018年苏拉威西岛地表覆盖信息,并从结构及类型变化、时空变化、动态变化3个角度分析苏拉威西岛地表覆盖的变化特征及变化驱动因素。研究表明:GEE在对海岛进行影像处理和地表覆盖信息提取方面具有处理数据量大、效率高的优势。随机森林分类精度较高,研究区2000年、2015年和2018年分类总体精度分别为91%、88%、85%,Kappa系数分别为0.89、0.86、0.82。变化监测研究发现苏拉威西岛的主要地类为林地和耕地,两者占整个海岛地表面积的85%以上。在2000~2018年期间受移民计划以及禁止非法砍伐政策的影响, 林地面积呈先减(减少7 982.29 km2)后增(增加9 079.17 km2)的态势,耕地面积不断萎缩(共减少14 267.35 km2)且主要流向林地、草地和人造地表。人造地表的变化最为活跃且其面积显著增加,主要驱动力为人口的迁移以及经济的发展。研究结果为生态型海岛的土地资源合理开发及应用、生态环境的保护、气候与环境变化等研究奠定了基础。

关键词: 海岛土地利用/土地覆盖随机森林变化监测及分析GEE    
Abstract:

With the increasing importance of the ocean in the context of country's politics, economy, and resources; the Land Use Land Cover (LULC) related research of island has received growing attention, which is significant to the bio-physical and socio-economic development for island. The Sulawesi island has been declared as globally important conservation area. Considering the problem of slow extraction and low efficiency of traditional large-scale land cover, this study took the Sulawesi as study area and carried out LULC change analysis based on the Google Earth Engine (GEE) platform. In this study Landsat TM/OLI images were used. After pre-processing all of the Landsat image, signature classes were selected based on the GEE platform. Finally, Random Forest (RF) algorithm was used to extract all of the land cover types of Sulawesi island from 2000 to 2018. In this study, change characteristics of land cover were analyzed from three aspects including structure and type change, spatio-temporal change, and dynamic change. This study observes the efficiency of GEE platform in regional scale data processing and surface cover extraction. The overall classification accuracy of the study in 2000, 2015 and 2018 are 91%, 88% and 85%, and Kappa coefficients are 0.89, 0.86 and 0.82, respectively. The LULC change analysis of this study shows that the main land cover types of Sulawesi Island are forest and cultivated land comprises more than 85% of the entire island. During the time period of 2000~2018, the forest land of Sulawesi was reduced first (reduced by 7 982.29 km2) and then increased (increased by 9 079.17 km2) due to the resettlement plan and the prohibiting of illegal logging while. The area of cultivated land was decreased (total reduced by 14 267.35 km2) converting to forest land, grassland, and artificial surface. Artificial surface has been increased significantly, mainly due to population migration and economic development in the island. The findings of this study will play an important role in guiding policy for regional development as well as resource management and environmental protection of Sulawesi island.

Key words: Island    Land Use/Land Cover    Random forest    Change monitoring and analysis    GEE
收稿日期: 2020-02-26 出版日期: 2021-04-13
ZTFLH:  P237  
基金资助: 中国科学院战略性先导科技专项(A类)(XDA13020506);国家自然科学基金项目(41771392)
通讯作者: 张丽     E-mail: futianmeng19@mails.ucas.ac.cn;zhangli@aircas.ac.cn
作者简介: 付甜梦(1995-),女,河南洛阳人,硕士研究生,主要从事生态遥感研究。E?mail:futianmeng19@mails.ucas.ac.cn
服务  
把本文推荐给朋友
加入引用管理器
E-mail Alert
RSS
作者相关文章  
付甜梦
张丽
陈博伟
闫敏

引用本文:

付甜梦,张丽,陈博伟,闫敏. 基于GEE平台的海岛地表覆盖提取及变化监测—以苏拉威西岛为例[J]. 遥感技术与应用, 2021, 36(1): 55-64.

Tianmeng Fu,Li Zhang,Bowei Chen,Min Yan. Monitoring of Land Cover Change based on Google Earth Engine Platform: A Case Study of Sulawesi Island. Remote Sensing Technology and Application, 2021, 36(1): 55-64.

链接本文:

http://www.rsta.ac.cn/CN/10.11873/j.issn.1004-0323.2021.1.0055        http://www.rsta.ac.cn/CN/Y2021/V36/I1/55

图1  苏拉威西区位图
地类特征

地物

类型

解译

标志

目视解译

人造

地表

人为活动造成的人工覆盖区域,如建筑(城市,城镇,交通),开采(露天采矿和采石场)或废物处理
裸地由于人类活动而没有人工掩护的区域,包括裸岩区,沙滩和沙漠。在影像中呈白色或黄色
水体包括天然水体和人工水体,如湖泊、河流、水库、水塘等。在影像中呈蓝黑色,纹理光滑
湿地在纯陆地系统和水生系统之间过渡的区域,为水生或定期被水淹的主要植被区域。主要由生长茂密的红树林组成
草地生长草本和灌木植物为主的土地
林地生长天然森林的土地,主要分布在山区
耕地指的是天然植被被移除或修改的地区,并被其他类型的人为来源的植被覆盖所取代,包括无植被覆盖的裸露耕地、果园以及种植园
表1  地表覆盖类型特征描述
指标含义公式注释
单一地表覆盖动态度某一地类在某时段内变化的幅度K=LUb-LUaLUc×1T×100%LUa:a时间某地类面积;LUb:b时间某地类面积;T:a、b时间间隔;ΔLUi-j:地类i转为非i地类的面积;n:地类的总数;Dab:从a时间到b时间,某一地类转入面积;Cab:从a时间到b时间,某一地类转出面积
综合地表覆盖动态度某时段内地表覆盖的数量变化程度LC=inΔLUi-j2inLUt×1T×100%
地表覆盖开发度单位时间内某一地类实际新开发的程度LUD=DabLUa×1T×100%
地表覆盖耗减度单位时间内某一种地类实际被消耗的程度

LUD=CabLUa×1T×100%

表2  地表覆盖动态变化指标描述
2000年2015年2018年
地类用户精度/%制图精度/%用户精度/%制图精度/%用户精度/%制图精度/%
人造地表82.3590.3287.1891.8972.7382.76
裸地93.3387.510010085.0080.95
水体86.6796.3093.9488.5784.3896.43
湿地96.0096.0095.8395.8396.30100
草地86.3686.3668.0077.2762.9680.95
林地91.4996.4895.3591.1195.4597.67
耕地96.3086.6782.1480.7093.3362.22
总体精度/%90.7588.0984.98
Kappa系数0.890.860.82
表3  苏拉威西岛各地表覆盖类型精度评价
图2  苏拉威西岛地表覆盖空间分布图
图3  苏拉威西岛各地类面积及变化情况图
2000年
地类人造地表裸地水体湿地草地林地耕地总计
2015年人造地表390.9020.18165.7123.64236.9286.181 612.752 536.27
裸地25.426.6615.532.2828.3813.88114.66206.81
水体54.8921.212 701.42146.939.0544.25776.663 754.41
湿地4.7317.76117.49561.561.261 227.08627.202 557.08
草地248.9830.94122.5632.311 289.363 224.445 242.3810 190.98
林地165.25348.71935.26260.55126.8692 121.854 591.5198 549.99
耕地613.4635.68780.93596.10805.909814.6340 050.4452 697.14
合计1 503.62481.164 838.891 623.372 497.74106 532.2853 015.61
变化面积1 112.72474.502 137.471 061.811 208.3814 410.4312 965.17
增(减)1 032.65-274.34-1 084.48933.717 693.24-7 982.29-318.47
表4  2000~2015年苏拉威西岛地表覆盖转移矩阵 (km2)
2015年
地类人造地表裸地水体湿地草地林地耕地总计
2018年人造地表1 281.7185.13108.398.61524.7666.901 772.473 847.96
裸地152.2643.7141.159.95121.5565.68156.07590.36
水体156.8211.623 337.2155.2233.8671.57835.524 501.82
湿地3.850.1864.32697.365.7032.93221.831 026.16
草地394.9530.7512.0110.445 763.751 517.756 419.3014148.95
林地8.962.2018.181 400.021 281.8595 160.699 757.23107 629.13
耕地537.7333.22173.15375.482 459.521 634.4433 534.7538 748.29
合计2 536.27206.813 754.412 557.0810 190.9898 549.9652 697.17
变化面积1 254.56163.10417.201 859.724 427.233 389.2719 162.42
增(减)1 311.69383.55747.41-1 530.923 957.979 079.17-13 948.88
表5  2015~2018年苏拉威西岛地表覆盖转移矩阵 (km2)
图4  2000~2018年苏拉威西岛地表覆盖类型转移图
图5  苏拉威西岛林地变化监测图
图6  苏拉威西岛耕地变化监测图
2000~2015年2015~2018年2000~2018年
指标KLUDLUCKLUDLUCK
人造地表/%4.589.514.9317.2433.7316.498.66
裸地/%-3.802.516.5761.8288.1126.291.26
水体/%-1.495.102.946.6410.343.70-0.39
湿地/%3.8310.484.36-19.964.2924.24-2.04
草地/%20.5326.543.2312.9527.4314.4825.92
林地/%-0.506.160.903.074.221.150.06
耕地/%-0.046.551.63-8.823.3012.12-1.50
表6  苏拉威西岛地表覆盖动态变化指标
1 Li Yaping, Yan Qin, Liu Zhengjun, et al. Extraction and Variation Analysis of Zhaoshu Isle Land Cover based on High Resolution Remote Sensing Images [J]. Journal of Tropical Oceanography, 2016, 35(2): 93-101.
1 李亚萍, 燕琴,刘正军,等. 基于高分辨率遥感影像的西沙赵述岛地表覆盖提取与演变分析[J]. 热带海洋学报, 2016, 35(2): 93-101.
2 LI Yaping. Change Detection and Analysis of Surface Coverage on Island and Reefs Via High Resolution Remote Sensing Image[D].Lanzhou: Lanzhou Jiaotong University, 2015.
2 李亚萍. 高分辨率遥感影像海岛礁地表覆盖变化检测与分析[D]. 兰州:兰州交通大学, 2015.
3 Zhang Linlin. Analysis Change on the Surface Coverage of Island based on Object-Oriented: Take Tongpan Island as Example[D]. Hangzhou: Hangzhou Normal University, 2018.张琳琳. 基于面向对象的海岛地表覆盖变化分析——以铜盘岛为例[D].杭州:杭州师范大学,2018.
4 Chen J, Pan D, Mao Z, et al. Land-cover Reconstruction and Change Analysis Using Multisource Remotely Sensed Imageries in Zhoushan Islands Since 1970[J]. Journal of Coastal Research, 2013, 30(2): 272-282. doi: 10.2112/JCOASTRES-D-13-00027.1.
doi: 10.2112/JCOASTRES-D-13-00027.1
5 Mei A, Manzo C, Fontinovo G, et al. Assessment of Land Cover Changes in Lampedusa Island(Italy) Using Landsat TM and OLI Data[J]. Journal of African Earth Sciences, 2016, 122: 15-24. doi:10.1016/j.jafrearsci.2015.05.014.
doi: 10.1016/j.jafrearsci.2015.05.014
6 Mageswaran T,Sachithanandam V, Sridhar R,et al. Mapping and Monitoring of Land Use/land Cover Changes in Neil Island(south Andaman) Using Geospatial Approaches[J].Indian Journal of Geo-Marine Sciences,2015,44(11):1762-1768.
7 Kumar L, Ghosh M K. Land Cover Change Detection of Hatiya Island,Bangladesh,Using Remote Sensing Techniques[J]. Journal of Applied Remote Sensing, 2012, 6(1): 063608. doi: 10.1117/1.jrs.6.063608.
doi: 10.1117/1.jrs.6.063608
8 Myers N, Mittermeier R A, Mittermeier C G, et al. Biodiversity Hotspots for Conservation Priorities[J]. Nature, 2000, 403(6772): 853. doi: 10.1038/35002501.
doi: 10.1038/35002501
9 Luo Yong, Jiang Ying, Dong Wenjie, et al. Climate Change and Sea Level Rise and Their Impacts on the Damage of Tsunami[J].Science & Technology Review. 2005, 23(0503): 41-44.罗勇, 江滢, 董文杰, 等. 气候变化与海平面上升及其对海啸灾害的影响[J]. 科技导报, 2005, 23(0503): 41-44.
10 Diffenbaugh N S, Singh D, Mankin J S, et al. Quantifying the Influence of Global Warming on Unprecedented Extreme Climate Events[J]. Proceedings of the National Academy of Sciences,2017,114(19):4881-4886. doi:10.1073/pnas.161808 2114.
doi: 10.1073/pnas.161808 2114
11 Cannon C H, Summers M, Harting J R, et al. Developing Conservation Priorities based on Forest Type, Condition, and Threats in a Poorly Known Ecoregion: Sulawesi, Indonesia [J]. Biotropica, 2007, 39(6): 747-759. doi: 10.1111/j.1744-7429.2007.00323.x.
doi: 10.1111/j.1744-7429.2007.00323.x
12 Xu Hanzeyu, Liu Chong, Wang Junbang, et al. Study on Extraction of Citrus Orchard in Gannan Region based on Google Earth Engine Platform [J]. Journal of Geo-information Science, 2018, 20(3):396-404.
12 徐晗泽宇, 刘冲, 王军邦,等. Google Earth Engine平台支持下的赣南柑橘果园遥感提取研究[J]. 地球信息科学学报, 2018, 20(3):396-404.
13 Liu Kai, Peng Liheng, Li Xiang, et al. Monitoring the Inter-annual Change of Mangrove based on the Google Earth Engine [J]. Journal of Geo-information Science, 2019,21(5):731-739.
13 刘凯, 彭力恒, 李想, 谭敏, 王树功. 基于Google Earth Engine的红树林年际变化监测研究[J]. 地球信息科学学报, 2019, 21(5): 731-739.
14 Di Gregorio A, Jansen L J M. Land Cover Classification System(LCCS): Classification Concepts and User Manual. Environment and Natural Resources Service[R]. GCP/RAF/287/ITA Africover—East Africa Project and Soil Resources, Management and Conservation Service, Food and Agriculture Organization, 2000.
15 Erasmi S, Twele A. Regional Land Cover Mapping in the Humid Tropics Using Combined Optical and SAR Satellite Data—A Case Study from Central Sulawesi, Indonesia[J]. International Journal of Remote Sensing, 2009, 30(10): 2465-2478. doi: 10.1080/01431160802552728.
doi: 10.1080/01431160802552728
16 Breiman L. Random Forests[J]. Machine Learning, 2001, 45(1): 5-32. doi: 10.1023/A:1010933404324.
doi: 10.1023/A:1010933404324
17 Gislason P O, Benediktsson J A, Sveinsson J R. Random Forests for Land Cover Classification[J]. Pattern Recognition Letters,2006,27(4):294-300. doi:10.1016/j.patrec.2005.08.011.
doi: 10.1016/j.patrec.2005.08.011
18 Waske B, Braun M. Classifier Ensembles for Land Cover Mapping Using Multitemporal SAR Imagery[J]. ISPRS Journal of Photogrammetry and Remote Sensing, 2009, 64(5):450-457. doi:10.1016/j.isprsjprs.2009.01.003.
doi: 10.1016/j.isprsjprs.2009.01.003
19 He Zhaoxin, Zhang Miao, Wu Bingfang, et al. Extraction of Summer Crop in Jiangsu based on Google Earth Engine [J]. Journal of Geo-information Science, 2019,21(5):752-766.
19 何昭欣,张淼,吴炳方,等. Google Earth Engine支持下的江苏省夏收作物遥感提取[J].地球信息科学学报,2019,21(5):752-766.
20 Hou Xiyong, Di Xianghong, Hou Wan, et al. Accuracy Evaluation of Land Use Mapping Using Remote Sensing Techniques in Coastal Zone of China[J]. Journal of Geo-information Science, 2018,20(10):1478-1488.
20 侯西勇,邸向红,侯婉,等.中国海岸带土地利用遥感制图及精度评价[J].地球信息科学学报,2018,20(10):1478-1488.
21 Wahyuni R. Forest Change Analysis Using OBIA Approach and Supervised Classification a Case Study: Kolaka District, South East Sulawesi[C]∥2017 International Conference on Advanced Computer Science and Information Systems, ICACSIS2017. doi:10.1109/ICACSIS.2017.8355019.
doi: 10.1109/ICACSIS.2017.8355019
22 Du Guoming, Kuang Wenhui, Meng Fanhao, et al. Spatiotemporal Pattern and Driving Forces of Land Use/cover Change in Brazil[J].Progress in Geography,2015,34(1):73-82.
22 杜国明,匡文慧,孟凡浩,等.巴西土地利用/覆盖变化时空格局及驱动因素[J].地理科学进展,2015,34(1): 73-82.
23 Shen Yanqing. An Analysis on Indonesia's Transmigration Program(1905~2000)[J]. Crossroads: Southeast Asian Studies, 2010(9):20-24.沈燕清.印度尼西亚国内移民计划浅析(1905~2000年)[J].东南亚纵横,2010(9):20-24.
24 Kim Y. Facts and Fictions of 'Transmigrasi' in Indonesia[J]. The Journal of Asian Studies, 2019, 22(1):75-100. doi: 10.21740/jas.2019.02.22.1.75.
doi: 10.21740/jas.2019.02.22.1.75
25 Liu Xinlai, Li Zhen. Law Enforcement Measures of Indonesian Forest Police Against Illegal Logging and Relevant Problems[J]. World Forestry Research,2019,32(4):82-85.
25 刘心来,李震.印度尼西亚森林警察打击非法采伐的执法对策及存在问题分析[J].世界林业研究,2019,32(4):82-85.
26 Jiang Fengping,Lu Wenming,Sun Rui,et al.Analysis of Illegal Logging in Indonesia[J]. World Forestry Research,2013,26(3):79-82.
26 姜凤萍,陆文明,孙睿,等.印度尼西亚木材非法采伐现状分析[J].世界林业研究,2013,26(3):79-82.
[1] 王崇阳,田昕. 基于GF⁃1 PMS数据的森林覆盖变化检测[J]. 遥感技术与应用, 2021, 36(1): 208-216.
[2] 马腾耀,肖鹏峰,张学良,马威,郭金金. 基于特征优选的GF-3全极化数据积雪识别[J]. 遥感技术与应用, 2020, 35(6): 1292-1302.
[3] 王一帆,徐涵秋. 基于客观阈值与随机森林Gini指标的水体遥感指数对比[J]. 遥感技术与应用, 2020, 35(5): 1089-1098.
[4] 李萌,年雁云,边瑞,白艳萍,马金辉. 基于多源遥感影像的青海云杉和祁连圆柏分类[J]. 遥感技术与应用, 2020, 35(4): 855-863.
[5] 张坤,刘乃文,高帅,赵书慧. 数据驱动的植被总初级生产力估算方法研究[J]. 遥感技术与应用, 2020, 35(4): 943-949.
[6] 李净,温松楠. 基于3种机器学习法的太阳辐射模拟研究[J]. 遥感技术与应用, 2020, 35(3): 615-622.
[7] 董超,赵庚星. 时序数据集构建质量对土地覆盖分类精度的影响研究[J]. 遥感技术与应用, 2020, 35(3): 558-566.
[8] 柴旭荣,李明,周义,王金风,田庆春. 影像的土地覆被快速分类[J]. 遥感技术与应用, 2020, 35(2): 315-325.
[9] 唐廷元,付波霖,何素云,娄佩卿,闭璐. 基于GF-1和Sentinel-1A的漓江流域典型地物信息提取[J]. 遥感技术与应用, 2020, 35(2): 448-457.
[10] 刘培,余志远,马威,韩瑞梅,陈正超,王涵,杨磊库. 基于地形信息的Landsat与Radarsat-2遥感数据协同分类研究[J]. 遥感技术与应用, 2019, 34(6): 1269-1275.
[11] 李哲,张沁雨,彭道黎. 基于高分二号遥感影像的树种分类方法[J]. 遥感技术与应用, 2019, 34(5): 970-982.
[12] 谷晓天, 高小红, 马慧娟, 史飞飞, 刘雪梅, 曹晓敏. 复杂地形区土地利用/土地覆被分类机器学习方法比较研究[J]. 遥感技术与应用, 2019, 34(1): 57-67.
[13] 刘立, 刘勇. 基于优化分割分类层次体系的土地覆被分类制图方法探讨[J]. 遥感技术与应用, 2019, 34(1): 79-89.
[14] 皋厦, 申鑫, 代劲松, 曹林. 结合LiDAR单木分割和高光谱特征提取的城市森林树种分类[J]. 遥感技术与应用, 2018, 33(6): 1073-1083.
[15] 陈伟民,张凌,宋冬梅,王斌,丁亚雄,许明明,崔建勇. 基于AdaBoost改进随机森林的高光谱图像地物分类方法研究[J]. 遥感技术与应用, 2018, 33(4): 612-620.