1.College of Forestry,Northeast Forestry University,Harbin 150040,China
2.Key Laboratory of Sustainable Forest Ecosystem Management,Ministry of Education,Harbin 150040,China
Show less
文章历史+
Received
Published
2025-03-19
2025-11-15
Issue Date
2025-12-24
PDF (1227K)
摘要
土地是人类生活中不可或缺的部分,分析土地利用现状有助于深入透彻地理解环境情况和经济发展之间的关系,实现更合理的土地利用模式。预测未来的土地利用情况有助于提高土地资源的可持续性管理,同时对评估碳潜力提供科学依据。以黑龙江省为研究区域,对黑龙江省2000—2020年土地利用现状进行分析,并采用斑块生成土地利用变化模拟模型(patch-generating land use simulation,PLUS)耦合长短期记忆模型(long short-term memory,LSTM)的方法,模拟预测黑龙江省2030年的土地利用情况。结果表明,1)验证PLUS-LSTM模型的Kappa系数为0.878,6种地类(耕地、林地、草地、水域、建设用地、未利用地)模拟相对误差均低于15%,相比于传统模型来说精度较高,可以用来模拟黑龙江省2030年土地利用情况;2)与2020年相比,2030年黑龙江省林地、草地、水域和建设用地的面积都有所增加。其中,建设用地的变化率最高,为8.57%;林地面积增加2 584.26 km²,扩张区域主要在中部地区;草地的扩张区域主要在西南部。耕地和未利用地面积减少,其中,未利用地变化最大,变化率为29.68%。
Abstract
Land is an indispensable part of human life. The analysis of land use status is helpful to deeply understand the relationship between environmental conditions and economic development, and to achieve a more reasonable land use model. Predicting future land use will help improve the sustainable management of land resources and provide a scientific basis for assessing carbon potential. Taking Heilongjiang Province as the research area, the current situation of land use in Heilongjiang Province from 2000 to 2020 was analyzed, and the patch-generating land use simulation (PLUS) model coupled with the long short-term memory (LSTM) model was adopted to simulate and predict the land use situation in Heilongjiang Province in 2030. The results showed that: 1) The Kappa coefficient for verifying the PLUS-LSTM model was 0.878. The relative simulation errors of the six land types (cultivated land, forest land, grassland, water area, construction land, and unused land) were all less than 15%. Compared with the traditional model, it had higher accuracy and can be used to simulate the land use situation in Heilongjiang Province in 2030. 2) Compared with 2020, the area of forest land, grassland, water area, and construction land in Heilongjiang Province would increase in 2030. Among them, the change rate of construction land was the highest, 8.57%; the area of forest land increased by 2 584.26 km², mainly in the central region; the expansion of grassland was mainly in the southwest. The area of cultivated land and unused land decreased, and the unused land changed the most, with a change rate of 29.68%.
土地不仅是人类生存和发展的核心基础,也为生物多样性和生态系统服务提供了重要保障[1]。土地利用覆盖变化(land use and cover change,LUCC)是人类活动影响全球环境变化的重要指标,对预测全球气候变化过程至关重要。土地利用变化趋势也成为未来陆地生态系统碳汇动态变化研究的重要参考因素。因此,如何深入理解其动态变化机制,准确估计模拟土地利用变化的时空特征,已成为当前学者研究的热点[2]。
早期的土地利用变化预测主要采用马尔可夫链模型(Markov)[3]、系统动力学模型(system dynamics,SD)[4]和Logistic回归(logistic regression,LR)模型[5]等,主要用于数量方面的预测,但是对于空间方面的预测难以实现,元细胞自动机模型(cellular automata,CA)[6]等主要用于空间方面的预测。但是土地利用变化的过程不仅涉及到了数量方面的变化,还包含了空间尺度的改变,因此越来越多的研究者将数量与空间方面的模型耦合起来进行土地利用现状模拟。通过土地利用转换及其效应模型(conversion of land use and its effect at small region extent,CLUE-S) [7]、CA-Markov模型[8]和未来土地利用模拟模型(future land-use simulation,FLUS)[9]等实现,但对于空间方面的模拟精度较低,同时近些年来多情景土地利用模拟需求增加,这些模型都有一定的局限性。相比之下,斑块生成土地利用变化模拟模型(patch-generating land use simulation,PLUS)于2021年提出,基于栅格数据,挖掘土地扩张的驱动因素来模拟预测未来土地利用情况的CA模型[10-12]。相较于CA模型而言,PLUS模型在空间上有自己独特的方法,模拟的精度更高,同时可以对不同情景的土地利用情况进行模拟,表现更好,因此越来越多的人开始使用PLUS模型进行研究[12]。周道媛等[13]利用PLUS模型模拟预测未来土地利用变化,进一步对生态系统服务价值进行评估。席梅竹等[14]利用PLUS模型对滹沱河流域山区段的生态格局进行了模拟,结果表明,相比于CA-Markov模型、FLUS-Markov模型而言,PLUS模型精度更高,同时可以对多情景土地利用情况进行模拟,效果更好。虽然PLUS模型对于空间尺度以及多情景预测具有不错的效果,但是在生态系统服务变化、生态保护以及碳储量评估方面也有不足,因此更多的学者将PLUS模型与其他模型耦合起来进行土地利用现状模拟以及生态系统服务的评估。王想等[15]利用PLUS-InVEST(integrated valuation of ecosystem services and trade-offs,InVEST)耦合模型对延庆区土地利用以及碳储量模拟进行预测,并提出生态优化方法,对推进延庆区可持续发展具有重要作用。
中国土地覆盖数据集(China land cover dataset,CLCD)是武汉大学利用谷歌地球引擎(google earth engine,GEE)基础设施创建的首个年度中国土地覆盖数据集。该数据集的准确率为79.31%,总体高于全球10 m地表覆盖数据(finer resolution observation and monitoring of global land cover,FROM-GLC)、30 m全球地表覆盖数据(GlobeLand30)、MODIS三级数据土地覆盖类型产品(moderate resolution imaging spectroradiometer,MCD12Q1)和欧航局300 m土地利用数据(ESA CCI land cover,ESACCI_LC)产品精度,具有较高的时空一致性[22]。研究将土地利用类型进行重分类,分为耕地、林地、草地、水域、建设用地以及未利用地6大类。耕地指用来种植农作物的土地;林地指被森林和林木覆盖的土地,包括天然林和人工林;草地是覆盖有草本植物的土地,包括天然草原和人工草地;水域包括河流、水库和湿地等水体;建设用地指用于建筑和基础设施的土地,包括住宅、商业、工业和公共设施用地等;未利用地指尚未开发或使用的土地,包括荒地、闲置地、裸地、沼泽区和自然保护区等。
JIAG, SHEVLIAKOVAE, ARTAXOP,et al.Land–climate interactions[M]//Special Report on Climate Change and Land: An IPCC Special Report on Climate Change,Desertification,Land Degradation,Sustainable Land Management,Food Security,and Greenhouse Gas Fluxes in Terrestrial Ecosystems.IPCC,2019:133-206.
LIUJ Y, NINGJ, KUANGW H,et al.Spatio-temporal patterns and characteristics of land-use change in China during 2010-2015[J].Acta Geographica Sinica,2018,73(5):789-802.
LAOG D.Analysis of the temporal and spatial evolution and driving force of the tidal flat in the Shanghai section of the north shore of Hangzhou Bay based on the Logistic-CA-Markovmodel[D].Shanghai:Shanghai Ocean University,2021.
MAS L, HUANGJ L.Multi-scenario simulation of low carbon land use in Changsha City based on SD-FLUS model[C]//China Association of Trade in Services,Chinese Society for Geodesy,Photogrammetry and Cartography,Chinese National Committee for Remote Sensing.Proceedings of the 16th Smart City Conference.Hunan Normal University,2023:191-206.
GUS Y.Driving force analysis of land use evolution in Nanchang city based on Logistic regression model[J].Rural Science and Technology,2023,14(17):114-117.
DENGY W, LING, ZENGS P.Simulation research on temporal and spatial changes of land use in Tianjin municipality based on CA-Markov coupling model[J].Journal of Tianjin Chengjian University,2021,27(6):387-395.
GUOH W, SUNX Y, LIANL S,et al.Response of water yield function of ecosystem to land use change in Nansi Lake Basin based on CLUE-S model and InVEST model[J].Chinese Journal of applied ecology,2016,27(9):2899-2906.
ZHUS H, SHUB R, MAX D,et al.The delimitation of urban growth boundary based on the idea of ‘anti-planning’ and FLUS Model:A case study of Jiawang District,Xuzhou City[J].Geography and Geo-Information Science,2017,33(5):80-86,127.
LEIX, HAIX Q.Assessing economic value of carbon storage and land use changes based on the coupled PLUS-InVEST model in Lanzhou City[J].Scientia Geographica Sinica,2025,45(2):339-348.
GUOW L, XUL Y, JIAJ A,et al.Predicting multi-scenario land use changes and soil erosion in the Huaihe river basin based on coupled PLUS-CSLE model[J].Journal of Soil and Water Conservation,2024,38(6):234-243,252.
[22]
LIANGX, GUANQ, CLARKEK C,et al.Understanding the drivers of sustainable land expansion using a patch-generating land use simulation (PLUS) model:A case study in Wuhan,China[J].Computers,Environment and Urban Systems,2021,85:101569.
ZHOUD Y, WANGJ, ZHOUF J.Evaluation of land use change and ecosystem service value based on multi-scenario analysis of PLUS model--with Fuzhou City as an example[J].Journal of Gansu Agricultural University,2023,58(3):198-209.
XIM Z, ZHAOZ Q, ZHAOY L.Multi-scenario prediction of the spatial and temporal patterns of ecological carrying capacity in the mountain section of Hutuo River basin in 2030 based on the PLUS model[J].Journal of Agricultural Resources and Environment,2024,41(3):518-529.
WANGX, WANGC Y, LYUF N,et al.Temporal and spatial variations of carbon storage and carbon sink improvement strategy at the district and county level based on PLUS-InVEST model:Taking Yanqing District as an example[J].Chinese Journal of Applied Ecology,2023,34(12):3373-3384.
[29]
JIAX, KHANDELWALA, NAYAKG,et al.Incremental dual-memory LSTM in land cover prediction[C]//Proceedings of the 23rd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining,2017:867-876.
[30]
WANGH, ZHAOX, ZHANGX,et al.Long time series land cover classification in China from 1982 to 2015 based on Bi-LSTM deep learning[J].Remote Sensing,2019,11(14):1639.
GONGS X, ZHANGY H, LIY H.Spatio-temporal variation and prediction of carbon storage in Beijing-Tianjin-Hebei region-A PLUS-InVEST model approach[J].Journal of Arid Land Resources and Environment,2023,37(6):20-28.
[37]
YANGJ, HUANGX.The 30 m annual land cover dataset and its dynamics in China from 1990 to 2019[J].Earth System Science Data,2021,13(8):3907-3925.
[38]
李国珍.基于FLUS模型的深圳市土地利用变化与模拟研究[D].武汉:武汉大学,2018.
[39]
LIG Z.Land use change and simulation in Shenzhen based on FLUS model[D].Wuhan:Wuhan University,2018.
XUEJ C, TAOZ M.Analysis on the difference and dynamic change of land use degree in different districts of Tianjin[J].Co-Operative Economy & Science,2019(8):8-11.
MAOH X, JIAK L, GAOX W,et al.Spatiotemporal pattern of land use change in Yinchuan Plain from 1980 to 2018[J].Science Technology and Engineering,2020,20(20):8008-8018.
LEIX, WANGQ N, DENGJ L,et al.Land use change and spatio-temporal simulation analysis of Meishan City based on FLUS model[J].Journal of Sichuan Normal University (Natural Science),2022,45(6):806-816.
ZHAOJ Y, CUIL, WANGJ,et al.Multi-scenario simulation of urban green space development and construction Timeline based on PLUS model analysis:A case study of the central Zhanjiang City[J].Acta Ecologica Sinica,2023,43(15):6307-6320.
MAR, FUY, LIUM S.Research on land use change prediction and landscape pattern evolution based on CA-MARKOV model:A case study of Chengdu City[J].Journal of Green Science and Technology,2023,25(22):245-252.
WANGX Y, TENGH.Analysis of spatio-temporal pattern evolution and driving factors of land use in Heilongjiang Province[J].Geomatics & Spatial Information Technology,2024,47(S1):129-132,139.