Land cover classification of Shangri-La by using Landsat data based on NDVI and canopy closure time-series characteristics considering fine classification of forests and grasslands

Abstract

Fine-scale classification data of forests and grasslands are essential for ecological protection, natural resource management, and environmental monitoring. In extensive mountainous regions, strong spectral mixing between forests and grasslands complicates the accurate delineation of their boundaries, while the lack of effective characteristics limits the refinement of forest type classification. This study proposes a fine classification method utilizing Landsat TM/ETM + /OLI data (1991–2021) to construct long-term temporal characteristics. A characteristic set of NDVI time series (NDVI standard curve, NM, and NP) was developed to delineate forests and grasslands and identify their transition (forest–steppe ecotone) based on NDVI differences. Meanwhile, seasonal metrics—seasonal variation characteristics of canopy closure were then extracted to classify 5 forest types, and the distribution above the tree line was used to identify 2 grassland types. Distinct from forests or grasslands, 6 additional land cover types were mapped using Maximum Likelihood Classification (MLC) based on spectral characteristics, producing an integrated 14-class land cover map. The final classification achieved User’s Accuracy (UA) values of 94.85%, 87.80%, and 84.72% for forests, grasslands, and the forest–steppe ecotone, respectively. Compared with classifications using only NDVI standard curve or spectral characteristics, the proposed NDVI characteristic set improved accuracy by 4.93 and 11.92% for forests, 7.50 and 12.43% for grasslands, and 12.98 and 34.52% for the ecotone. All forest and grassland subtypes exceeded 80% accuracy. Compared with existing datasets, such as GlobeLand30, CLCD, and FROM-GLC (2017), the method more precisely captured the forest-steppe ecotone and enhanced the refinement of forest and grassland classification, improving forest classification accuracy by 8.07%, 6.87%, and 8.77%, and grassland accuracy by 7.36%, 25.96%, and 12.66%, respectively. The current study establishes a basis for investigating the physical attributes of the land cover classification model. This study offers novel concepts for developing efficient remote sensing classification characteristics for land cover.

Keywords

Multi-temporal classification / NDVI time series / Landsat / Canopy closure / Tree line extraction

References

[1] Belayneh Y, Ru G, Guadie A, Teffera ZL, Tsega M. Forest cover change and its driving forces in Fagita Lekoma District, Ethiopia. J for Res. 2020, 31(5): 1567-1582.

[2] Berra EF, Gaulton R. Remote sensing of temperate and boreal forest phenology: a review of progress, challenges and opportunities in the intercomparison of in-situ and satellite phenological metrics. For Ecol Manage. 2021, 480. 118663

[3] Bovolo F, Bruzzone L. The time variable in data fusion: a change detection perspective. IEEE Geosci Remote Sens Mag. 2015, 3(3): 8-26.

[4] Braun A. Retrieval of digital elevation models from Sentinel-1 radar data—open applications, techniques, and limitations. Open Geosci. 2021, 13(1): 532-569.

[5] Bruzzone L, Roli F, Serpico SB. An extension of the Jeffreys-Matusita distance to multiclass cases for feature selection. IEEE Trans Geosci Remote Sens. 1995, 33(6): 1318-1321.

[6] Chang DHS. The vegetation zonation of the Tibetan Plateau. Mt Res Dev. 1981, 1(1): 29.

[7] Chen X, Shi XL. Geoscience landscape division and tourism zonation in the mid-southern section of the Hengduan Mountains, eastern Qinghai-Tibet Plateau. J Mt Sci. 2018, 154894-917.

[8] Chen J, Chen J, Liao AP, Cao X, Chen LJ, Chen XH, He CY, Han G, Peng S, Lu M, Zhang WW, Tong XH, Mills J. Global land cover mapping at 30m resolution: a POK-based operational approach. ISPRS J Photogramm Remote Sens. 2015, 103: 7-27.

[9] Claverie M, Ju JC, Masek JG, Dungan JL, Vermote EF, Roger JC, Skakun SV, Justice C. The Harmonized Landsat and Sentinel-2 surface reflectance data set. Remote Sens Environ. 2018, 219: 145-161.

[10] Dallahi Y, Boujraf A, Meliho M, Orlando CA. Assessment of forest dieback on the Moroccan Central Plateau using spectral vegetation indices. J for Res. 2023, 34(3): 793-808.

[11] Ding MJ, Zhang YL, Liu LS, Zhang W, Wang ZF, Bai WQ. The relationship between NDVI and precipitation on the Tibetan Plateau. J Geogr Sci. 2007, 17(3): 259-268.

[12] Doughty CE, Keany JM, Wiebe BC, Rey-Sanchez C, Carter KR, Middleby KB, Cheesman AW, Goulden ML, da Rocha HR, Miller SD, Malhi Y, Fauset S, Gloor E, Slot M, Oliveras Menor I, Crous KY, Goldsmith GR, Fisher JB. Tropical forests are approaching critical temperature thresholds. Nature. 2023, 621(7977): 105-111.

[13] Falkowski MJ, Gessler PE, Morgan P, Hudak AT, Smith AMS. Characterizing and mapping forest fire fuels using ASTER imagery and gradient modeling. For Ecol Manag. 2005, 217(2–3): 129-146.

[14] Foushee J (2022) Multi-decadal analysis of remotely sensed vegetation change in Berea College Forest -seasonality of forest patterns using remote sensing. College of Arts & Sciences Senior Honors Theses. 27. https://ir.library.louisville.edu/honors/275

[15] Galford GL, Mustard JF, Melillo J, Gendrin A, Cerri CC, Cerri CEP. Wavelet analysis of MODIS time series to detect expansion and intensification of row-crop agriculture in Brazil. Remote Sens Environ. 2008, 112(2): 576-587.

[16] Gong P, Miller JR, Spanner M. Forest canopy closure from classification and spectral unmixing of scene components-multisensor evaluation of an open canopy. IEEE Trans Geosci Remote Sens. 1994, 32(5): 1067-1080.

[17] Gong P, Wang J, Yu L, Zhao YC, Zhao YY, Liang L, Niu ZG, Huang XM, Fu HH, Liu S, Li CC, Li XY, Fu W, Liu CX, Xu Y, Wang XY, Cheng Q, Hu LY, Yao WB, Zhang H, Zhu P, Zhao ZY, Zhang HY, Zheng YM, Ji LY, Zhang YW, Chen H, Yan A, Guo JH, Yu L, Wang L, Liu XJ, Shi TT, Zhu MH, Chen YL, Yang GW, Tang P, Xu B, Giri C, Clinton N, Zhu ZL, Chen J, Chen J. Finer resolution observation and monitoring of global land cover: first mapping results with Landsat TM and ETM+ data. Int J Remote Sens. 2013, 34(7): 2607-2654.

[18] Gong Z, Ge WY, Guo JQ, Liu JC. Satellite remote sensing of vegetation phenology: progress, challenges, and opportunities. ISPRS J Photogramm Remote Sens. 2024, 217: 149-164.

[19] Guo QH, Yu H, Cao YL, Zhang ZP (1997) The remote sensing study on the characteristics of forest-steppe ecotone. Acta Scientiarum Naturalium Universitatis Pekinensis. 35(4). https://link.cnki.net/doi/https://doi.org/10.13209/j.0479-8023.1999.082

[20] Han W, Feng RY, Wang LZ, Cheng YF. A semi-supervised generative framework with deep learning features for high-resolution remote sensing image scene classification. ISPRS J Photogramm Remote Sens. 2018, 145: 23-43.

[21] Harsch MA, Hulme PE, McGlone MS, Duncan RP. Are treelines advancing? A global meta-analysis of treeline response to climate warming. Ecol Lett. 2009, 12(10): 1040-1049.

[22] He YJ. Water and heat conditions seperately controlled inter-annual variation and growth trend of NDVI in the temperate grasslands in China. Acta Ecol Sin. 2022, 42(2): 766-777.

[23] Hong DF, Yokoya N, Xia GS, Chanussot J, Zhu XX. X-ModalNet: a semi-supervised deep cross-modal network for classification of remote sensing data. ISPRS J Photogramm Remote Sens. 2020, 167: 12-23.

[24] Huang XY, Yin YW, Feng LW, Tong XY, Zhang XX, Li JR, Tian F. A 10 m resolution land cover map of the Tibetan Plateau with detailed vegetation types. Earth Syst Sci Data. 2024, 16(7): 3307-3332.

[25] Huete AR. A soil-adjusted vegetation index (SAVI). Remote Sens Environ. 1988, 25(3): 295-309.

[26] Huete A, Didan K, Miura T, Rodriguez EP, Gao X, Ferreira LG. Overview of the radiometric and biophysical performance of the MODIS vegetation indices. Remote Sens Environ. 2002, 83(1–2): 195-213.

[27] Keogh E, Ratanamahatana CA. Exact indexing of dynamic time warping. Knowl Inf Syst. 2005, 7(3): 358-386.

[28] Koelemeijer IA, Ehrlén J, Jönsson M, De Frenne P, Berg P, Andersson J, Weibull H, Hylander K (2022) Interactive effects of drought and edge exposure on old-growth forest understory species. Landsc Ecol 37(7):1839–1853. https://doi.org/10.1007/s10980-022-01441-9

[29] Körner C. The use of ‘altitude’ in ecological research. Trends Ecol Evol. 2007, 22(11): 569-574.

[30] Kramer RD, Ishii HR, Carter KR, Miyazaki Y, Cavaleri MA, Araki MG, Azuma WA, Inoue Y, Hara C. Predicting effects of climate change on productivity and persistence of forest trees. Ecol Res. 2020, 35(4): 562-574.

[31] Kwenda C, Gwetu MV, Fonou-Dombeu JV (2023) Forest image classification based on deep learning and XGBoost algorithm. Computational Science – ICCS 2023.217–29. https://doi.org/10.1007/978-3-031-36027-5_16

[32] Li ZZ, Bi SD, Hao S, Cui YH. Aboveground biomass estimation in forests with random forest and Monte Carlo-based uncertainty analysis. Ecol Indic. 2022, 142. 109246

[33] Li Y, Huang B, Rust HW. Using statistical models to depict the response of multi-timescale drought to forest cover change across climate zones. Hydrol Earth Syst Sci. 2024, 28(2): 321-339.

[34] Lin SK. Introduction to remote sensing. Fifth edition. By James B. Campbell and Randolph H. Wynne, the GuilfordPress, 2011, 662 pages. ISBN 978-1-60918-176-5. Remote Sens. 2013, 5(1): 282-283.

[35] Liu B. Vertical patterns in plant diversity and their relations with environmental factors on the southern slope of the Tianshan Mountains (middle section) in Xinjiang (China). J Mt Sci. 2017, 14(4): 742-757.

[36] Liu LY, Zhang X, Gao Y, Chen XD, Shuai X, Mi J. Finer-resolution mapping of global land cover: recent developments, consistency analysis, and prospects. J Remote Sens. 2021, 2021. 2021/5289697

[37] Mahmoud SH, Gan TY. Impact of anthropogenic climate change and human activities on environment and ecosystem services in arid regions. Sci Total Environ. 2018, 633: 1329-1344.

[38] Martinuzzi F, Mahecha MD, Camps-Valls G, Montero D, Williams T, Mora K. Learning extreme vegetation response to climate drivers with recurrent neural networks. Nonlin Processes Geophys. 2024, 31(4): 535-557.

[39] Mathisen IE, Mikheeva A, Tutubalina OV, Aune S, Hofgaard A. Fifty years of tree line change in the Khibiny Mountains, Russia: advantages of combined remote sensing and dendroecological approaches. Appl Veg Sci. 2014, 17(1): 6-16.

[40] Mehmood K, Ahmad Anees S, Rehman A, Tariq A, Liu QJ, Muhammad S, Rabbi F, Pan SA, Hatamleh WA. Assessing forest cover changes and fragmentation in the Himalayan temperate region: implications for forest conservation and management. J Forestry Res. 2024, 35. 82

[41] Morley PJ, Donoghue DNM, Chen JC, Jump AS. Integrating remote sensing and demography for more efficient and effective assessment of changing mountain forest distribution. Ecol Inform. 2018, 43: 106-115.

[42] Myneni RB, Keeling CD, Tucker CJ, Asrar G, Nemani RR. Increased plant growth in the northern high latitudes from 1981 to 1991. Nature. 1997, 386(6626): 698-702.

[43] Nagai S, Saitoh TM, Kobayashi H, Ishihara M, Suzuki R, Motohka T, Nasahara KN, Muraoka H. In situ examination of the relationship between various vegetation indices and canopy phenology in an evergreen coniferous forest, Japan. Int J Remote Sens. 2012, 33(19): 6202-6214.

[44] Nedd R, Light K, Owens M, James N, Johnson E, Anandhi A. A synthesis of land use/land cover studies: definitions, classification systems, meta-studies, challenges and knowledge gaps on a global landscape. Land (Basel). 2021, 10(9. 994

[45] Nemani RR, Keeling CD, Hashimoto H, Jolly WM, Piper SC, Tucker CJ, Myneni RB, Running SW. Climate-driven increases in global terrestrial net primary production from 1982 to 1999. Science. 2003, 3005625): 1560-1563.

[46] Peng J, Tian L, Zhang ZM, Zhao Y, Green SM, Quine TA, Liu HY, Meersmans J. Distinguishing the impacts of land use and climate change on ecosystem services in a karst landscape in China. Ecosyst Serv. 2020, 46. 101199

[47] Pettorelli N, Vik JO, Mysterud A, Gaillard JM, Tucker CJ, Stenseth NC. Using the satellite-derived NDVI to assess ecological responses to environmental change. Trends Ecol Evol. 2005, 20(9): 503-510.

[48] Piao SL, Wang XH, Ciais P, Zhu B, Wang T, Liu J. Changes in satellite-derived vegetation growth trend in temperate and boreal Eurasia from 1982 to 2006. Glob Change Biol. 2011, 17(10): 3228-3239.

[49] Rifai SW, De Kauwe MG, Ukkola AM, Cernusak LA, Meir P, Medlyn BE, Pitman AJ. Thirty-eight years of CO₂ fertilization has outpaced growing aridity to drive greening of Australian woody ecosystems. Biogeosciences. 2022, 19(2): 491-515.

[50] Rota N, Canedoli C, Ferrè C, Ficetola GF, Guerrieri A, Padoa-Schioppa E. Evaluation of soil biodiversity in Alpine habitats through eDNA metabarcoding and relationships with environmental features. Forests. 2020, 11(7): 738.

[51] Sadian A, Shafizadeh-Moghadam H. The future of agricultural lands under the combined influence of shared socioeconomic pathways and urban expansion by 2050. Agric Syst. 2025, 224. 104234

[52] Santos Costa W, Fonseca LMG, Körting TS, do Nascimento Bendini H, de Cartaxo Mosto Souza R. Spatio-temporal segmentation applied to optical remote sensing image time series. IEEE Geosci Remote Sens Lett. 2018, 1581299-1303.

[53] Senf C, Leitão PJ, Pflugmacher D, van der Linden S, Hostert P. Mapping land cover in complex mediterranean landscapes using Landsat: improved classification accuracies from integrating multi-seasonal and synthetic imagery. Remote Sens Environ. 2015, 156: 527-536.

[54] Shafizadeh-Moghadam H, Khazaei M, Alavipanah SK, Weng QH. Google Earth Engine for large-scale land use and land cover mapping: an object-based classification approach using spectral, textural and topographical factors. Gisci Remote Sens. 2021, 58(6): 914-928.

[55] Sulla-Menashe D, Friedl MA, Krankina ON, Baccini A, Woodcock CE, Sibley A, Sun GQ, Kharuk V, Elsakov V. Hierarchical mapping of Northern Eurasian land cover using MODIS data. Remote Sens Environ. 2011, 115(2): 392-403.

[56] Sun C, Li JL, Liu YX, Liu YC, Liu RQ. Plant species classification in salt marshes using phenological parameters derived from Sentinel-2 pixel-differential time-series. Remote Sens Environ. 2021, 256. 112320

[57] Sun X, Tian Y, Lu WX, Wang PJ, Niu RG, Yu HF, Fu K. From single- to multi-modal remote sensing imagery interpretation: a survey and taxonomy. Sci China Inf Sci. 2023, 66(4. 140301

[58] Tang ZY, Wang ZH, Zheng CY, Fang JY. Biodiversity in China’s mountains. Front Ecol Environ. 2006, 47): 347-352.

[59] Tang LN, Gao LJ, Shi LY. Sustainable management and protection of ecosystems in Shangri-La County, Yunnan Province, China: introduction. Int J Sustain Dev World Ecol. 2015, 22(2): 99-102.

[60] Tucker CJ. Red and photographic infrared linear combinations for monitoring vegetation. Remote Sens Environ. 1979, 8(2): 127-150.

[61] Veloso A, Mermoz S, Bouvet A, Le Toan T, Planells M, Dejoux JF, Ceschia E. Understanding the temporal behavior of crops using Sentinel-1 and Sentinel-2-like data for agricultural applications. Remote Sens Environ. 2017, 199: 415-426.

[62] Wang R, Gamon J, Montgomery R, Townsend P, Zygielbaum A, Bitan KR, Tilman D, Cavender-Bares J. Seasonal variation in the NDVI–species richness relationship in a prairie grassland experiment (Cedar Creek). Remote Sens. 2016, 8(2): 128.

[63] Wang LJ, Wang JY, Liu ZZ, Zhu J, Qin F. Evaluation of a deep-learning model for multispectral remote sensing of land use and crop classification. Crop J. 2022, 1051435-1451.

[64] Wang Y, Albrecht CM, Ali Braham NA, Mou LC, Zhu XX. Self-supervised learning in remote sensing: a review. IEEE Geosci Remote Sens Mag. 2022, 10(4): 213-247.

[65] Wang RL, Wang M, Sun XF, Wang JB, Li GC. Enhancing forest-steppe ecotone mapping accuracy through synthetic ApertureRadar-optical remote sensing data fusion and object-based analysis. Photogramm Eng Remote Sensing. 2024, 90(7): 415-426.

[66] Wu YF (2021) Research on the estimation of forest parameters based on remote sensing image data: a case study of Chun’an County, Zhejiang Province. Master’s thesis, Agricultural Engineering and Information Technology, College of Environment and Resources, Zhejiang University. https://doi.org/10.27461/d.cnki.gzjdx.2021.000652

[67] Wu Y, Zhao R, Hu Q, Zhang YJ, Zhang K. Retrieving sub-canopy terrain from ICESat-2 data based on the RNR-DCM filtering and erroneous ground photons correction approach. Remote Sens. 2023, 15(15. 3904

[68] Wu, ZY, Zhu, YC, Jiang, HQ (1987) Vegetation of Yunnan. Science Press, Beijing, p 1–1024. (in Chinese)

[69] Wulder M. Optical remote-sensing techniques for the assessment of forest inventory and biophysical parameters. Prog Phys Geogr Earth Environ. 1998, 224): 449-476.

[70] Xing YJ, Chen MH, Dao JC, Lin LX, Chen CY, Chen YL, Wang ZT. Fine-root morphology of woody and herbaceous plants responds differently to altered precipitation: a meta-analysis. For Ecol Manag. 2024, 552. 121570

[71] Xu WY, Jin XB, Liu J, Yang XH, Ren J, Zhou YK. Analysis of spatio-temporal changes in forest biomass in China. J Forestry Res. 2022, 33(1): 261-278.

[72] Yan WY. Airborne lidar data artifacts: what we know thus far.. IEEE Geosci Remote Sens Mag. 2023, 11(3): 21-45.

[73] Yang J, Huang X. The 30 m annual land cover dataset and its dynamics in China from 1990 to 2019. Earth Syst Sci Data. 2021, 13(8): 3907-3925.

[74] Yun HW, Kim JR, Choi YS, Lin SY. Analyses of time series InSAR signatures for land cover classification: case studies over dense forestry areas with L-band SAR images.. Sensors. 2019, 19(12. 2830

[75] Zhang C, Zhang L, Zhang BYJ, Sun JQ, Dong SK, Wang XY, Li YX, Xu J, Chu WK, Dong YW, Wang P. Land cover classification in a mixed forest-grassland ecosystem using LResU-net and UAV imagery. J Forestry Res. 2022, 333): 923-936.

[76] Zhao Z, Wang JS, Wang LM, Rao X, Ran WJ, Xu CX. Monitoring and analysis of abandoned cropland in the Karst Plateau of eastern Yunnan, China based on Landsat time series images. Ecol Indic. 2023, 146. 109828

[77] Zhou LM, Tucker CJ, Kaufmann RK, Slayback D, Shabanov NV, Myneni RB. Variations in northern vegetation activity inferred from satellite data of vegetation index during 1981 to 1999. J Geophys Res. 2001, 106(D17): 20069-20083.

[78] Zhou ZL, Wu JN, Wang C, Wang JL, Cheng F. A synchronous acquisition method for dominant tree species and forest age in complex mountainous terrain through growth characteristics matching. IEEE Trans Geosci Remote Sens. 2025, 63: 4403518.

[79] Zhu Z, Woodcock CE, Olofsson P. Continuous monitoring of forest disturbance using all available Landsat imagery. Remote Sens Environ. 2012, 12275-91.

RIGHTS & PERMISSIONS

Northeast Forestry University