Dominant woody plant species recognition with a hierarchical model based on multimodal geospatial data for subtropical forests

Abstract

Since the launch of the Google Earth Engine (GEE) cloud platform in 2010, it has been widely used, leading to a wealth of valuable information. However, the potential of GEE for forest resource management has not been fully exploited. To extract dominant woody plant species, GEE combined Sentinel-1 (S1) and Sentinel-2 (S2) data with the addition of the National Forest Resources Inventory (NFRI) and topographic data, resulting in a 10 m resolution multimodal geospatial dataset for subtropical forests in southeast China. Spectral and texture features, red-edge bands, and vegetation indices of S1 and S2 data were computed. A hierarchical model obtained information on forest distribution and area and the dominant woody plant species. The results suggest that combining data sources from the S1 winter and S2 yearly ranges enhances accuracy in forest distribution and area extraction compared to using either data source independently. Similarly, for dominant woody species recognition, using S1 winter and S2 data across all four seasons was accurate. Including terrain factors and removing spatial correlation from NFRI sample points further improved the recognition accuracy. The optimal forest extraction achieved an overall accuracy (OA) of 97.4% and a map-level image classification efficacy (MICE) of 96.7%. OA and MICE were 83.6% and 80.7% for dominant species extraction, respectively. The high accuracy and efficacy values indicate that the hierarchical recognition model based on multimodal remote sensing data performed extremely well for extracting information about dominant woody plant species. Visualizing the results using the GEE application allows for an intuitive display of forest and species distribution, offering significant convenience for forest resource monitoring.

Keywords

Google Earth Engine / Sentinel / Forest resource inventory data / Dominant woody plant species / Subtropics / Model performance

References

[1]	Abera TA, Vuorinne I, Munyao M, Pellikka PKE, Heiskanen J (2022) Land cover map for multifunctional landscapes of Taita Taveta County, Kenya, based on sentinel-1 radar, sentinel-2 optical, and topoclimatic data. Data 7:36. https://doi.org/10.3390/data7030036
[2]	Adrian J, Sagan V, Maimaitijiang M (2021) Sentinel SAR-optical fusion for crop type mapping using deep learning and google earth engine. ISPRS J Photogramm 175:215–235. https://doi.org/10.1016/j.isprsjprs.2021.02.018
[3]	Amani M, Ghorbanian A, Ahmadi SA, Kakooei M, Moghimi A, Mirmazloumi SM, Moghaddam SHA, Mahdavi S, Ghahremanloo M, Parsian S, Wu QS, Brisco B (2020) Google earth engine cloud computing platform for remote sensing big data applications: a comprehensive review. IEEE J-STARS 13:5326–5350. https://doi.org/10.1109/JSTARS.2020.3021052
[4]	Amissah L, Mohren GMJ, Bongers F, Hawthorne WD, Poorter L (2014) Rainfall and temperature affect tree species distribution in Ghana. J Trop Ecol 30:435–446. https://doi.org/10.1017/S026646741400025X
[5]	Bhattarai R, Rahimzadeh-Bajgiran P, Weiskittel A, Meneghini A, MacLean DA (2021) Spruce budworm tree host species distribution and abundance mapping using multi-temporal Sentinel-1 and Sentinel-2 satellite imagery. ISPRS J Photogramm 172:28–40. https://doi.org/10.1016/j.isprsjprs.2020.11.023
[6]	Brown CF, Brumby SP, Guzder-Williams B, Birch T, Hyde SB, Mazzariello J, Czerwinski W, Pasquarella VJ, Haertel R, Ilyushchenko S, Schwehr K, Weisse M, Stolle F, Hanson C, Guinan O, Moore R, Tait AM (2022) Dynamic world, near real-time global 10 m land use land cover mapping. Sci Data 9:251. https://doi.org/10.1038/s41597-022-01307-4
[7]	Cheng K, Wang JL, Yan XR (2021) Mapping forest types in China with 10 m resolution based on spectral-spatial-temporal features. Remote Sens-Basel 13:973. https://doi.org/10.3390/rs13050973
[8]	Chiang SH, Valdez M (2019) Tree species classification by integrating satellite imagery and topographic variables using maximum entropy method in a Mongolian forest. Forests 10:961. https://doi.org/10.3390/f10110961
[9]	Conners RW, Trivedi MM, Harlow CA (1984) Segmentation of a high-resolution urban scene using texture operators. CVGIP 25:273–310. https://doi.org/10.1016/0734-189X(84)90197-X
[10]	DEM. NASA JPL (2020) NASADEM merged DEM global 1 arc second V001 [data set]. NASA EOSDIS Land Processes DAAC. https://doi.org/10.5067/MEaSUREs/NASADEM/NASADEM_HGT.001
[11]	Deur M, Gasparovic M, Balenovic I (2020) Tree species classification in mixed deciduous forests using very high spatial resolution satellite imagery and machine learning methods. Remote Sens-Basel 12:3926. https://doi.org/10.3390/rs12233926
[12]	Dostálová A, Hollaus M, Milenkovi? M, Wagner W (2016) Forest area derivation from sentinel-1 data. ISPRS Ann Photogramm Remote Sens Spat Inf Sci. 3(7):227–233
[13]	Dyderski MK, Pa? S, Frelich LE, Jagodziński AM (2018) How much does climate change threaten European forest tree species distributions? Glob Change Biol 24:1150–1163. https://doi.org/10.1111/gcb.13925
[14]	Fassnacht FE, Latifi H, Stereńczak K, Modzelewska A, Lefsky M, Waser LT, Straub C, Ghosh A (2016) Review of studies on tree species classification from remotely sensed data. Remote Sens Environ 186:64–87. https://doi.org/10.1016/j.rse.2016.08.013
[15]	Foga S, Scaramuzza PL, Guo S, Zhu Z, Dilley RD, Beckmann T, Schmidt GL, Dwyer JL, Joseph-Hughes M, Laue B (2017) Cloud detection algorithm comparison and validation for operational Landsat data products. Remote Sens Environ 194:379–390. https://doi.org/10.1016/j.rse.2017.03.026
[16]	Fricker GA, Ventura JD, Wolf JA, North MP, Davis FW, Franklin J (2019) A convolutional neural network classifier identifies tree species in mixed-conifer forest from hyperspectral imagery. Remote Sens-Basel 11:2326. https://doi.org/10.3390/rs11192326
[17]	Friedman JH (2001) Greedy function approximation: a gradient boosting machine. Ann Stat 29:1189–1232. https://doi.org/10.1214/aos/1013203451
[18]	Ghorbanian A, Zaghian S, Asiyabi RM, Amani M, Mohammadzadeh A, Jamali S (2021) Mangrove ecosystem mapping using sentinel-1 and sentinel-2 satellite images and random forest algorithm in Google Earth Engine. Remote Sens-Basel 13:2565. https://doi.org/10.3390/rs13132565
[19]	Gorelick N, Hancher M, Dixon M, Ilyushchenko S, Thau D, Moore R (2017) Google earth engine: planetary-scale geospatial analysis for everyone. Remote Sens Environ 202:18–27. https://doi.org/10.1016/j.rse.2017.06.031
[20]	Guo YT, Jiao WL, Long TF, Zhang XM (2022) Generation of the forest cover map of the Qinghai-Tibet plateau based on the multisource dataset and random forest. Front Environ Sci Eng 10:939151. https://doi.org/10.3389/fenvs.2022.939151
[21]	Hansen MC, Potapov PV, Moore R, Hancher M, Turubanova SA, Tyukavina A, Thau D, Stehman SV, Goetz SJ, Loveland LR, Kommareddy A, Egorov A, Chini L, Justice CO, Townshend J (2013) High-resolution global maps of 21st-century forest cover change. Science 342:850–853. https://doi.org/10.1126/science.1244693
[22]	Haralick RM, Shanmugam K, Dinstein I (1973) Textural features for image classification. IEEE Trans Syst Man Cybern SMC 3:610–621. https://doi.org/10.1109/TSMC.1973.4309314
[23]	He T, Zhou HK, Xu CY, Hu JG, Xue XY, Xu LC, Lou XW, Zeng K, Wang Q (2023) Deep learning in forest tree species classification using sentinel-2 on google earth engine: a case study of Qingyuan County. Sustain-Basel 15:2741. https://doi.org/10.3390/su15032741
[24]	Hemmerling J, Pflugmacher D, Hostert P (2021) Mapping temperate forest tree species using dense sentinel-2 time series. Remote Sens Environ 267:112743. https://doi.org/10.1016/j.rse.2021.112743
[25]	Immitzer M, Neuwirth M, Boeck S, Brenner H, Vuolo F, Atzberger C (2019) Optimal input features for tree species classification in central Europe based on multi-temporal Sentinel-2 data. Remote Sens-Basel 11:2599. https://doi.org/10.3390/rs11222599
[26]	Isbaex C, Margarida Coelho A (2021) The potential of sentinel-2 satellite images for land-cover/land-use and forest biomass estimation: a review. In: Forest biomass-from trees to energy, IntechOpen, London, pp 1–24. https://doi.org/10.5772/intechopen.93363
[27]	Jiang YF, Zhang L, Yan M, Qi JG, Fu TM, Fan SX, Chen BW (2021) High-resolution mangrove forests classification with machine learning using worldview and UAV hyperspectral data. Remote Sens-Basel 13:1529. https://doi.org/10.3390/rs13081529
[28]	Karra K, Kontgis C, Statman-Weil Z, Mazzariello JC, Mathis M, Brumby SP (2021) Global land use/land cover with sentinel 2 and deep learning. In: 2021 IEEE international geoscience and remote sensing symposium IGARSS. pp. 4704–4707. https://doi.org/10.1109/IGARSS47720.2021.9553499
[29]	Kumar L, Mutanga O (2018) Google earth engine applications since inception: usage, trends, and potential. Remote Sens-Basel 10:1509. https://doi.org/10.3390/rs10101509
[30]	Lechner AM, Foody GM, Boyd DS (2020) Applications in remote sensing to forest ecology and management. One Earth 2:405–412. https://doi.org/10.1016/j.oneear.2020.05.001
[31]	Li RN, Fang PF, Xu WH, Wang LG, Ou GL, Zhang WQ, Huang X (2022) Classifying forest types over a mountainous area in southwest China with landsat data composites and multiple environmental factors. Forests 13:135. https://doi.org/10.3390/f13010135
[32]	Lin YC, Shao JY, Shin SY, Saka Z, Joseph M, Manish R, Fei SL, Habib A (2022) Comparative analysis of multi-platform, multi-resolution, multi-temporal LiDAR data for forest inventory. Remote Sens-Basel 14:649. https://doi.org/10.3390/rs14030649
[33]	Luo C, Qi BS, Liu HJ, Guo D, Lu LP, Fu Q, Shao YQ (2021) Using time series sentinel-1 images for object-oriented crop classification in google earth engine. Remote Sens-Basel 13:561. https://doi.org/10.3390/rs13040561
[34]	Ma M, Liu J, Liu M, Zeng JC, Li YH (2021) Tree species classification based on sentinel-2 imagery and random forest classifier in the eastern regions of the Qilian mountains. Forests 12:1736. https://doi.org/10.3390/f12121736
[35]	Malenovsky Z, Rott H, Cihlar J, Schaepman ME, García-Santos G, Fernandes R, Berger M (2012) Sentinels for science: potential of sentinel-1, -2, and -3 missions for scientific observations of ocean, cryosphere, and land. Remote Sens Environ 120:91–101. https://doi.org/10.1016/j.rse.2011.09.026
[36]	Michalowska M, Rapinski J (2021) A review of tree species classification based on airborne LiDAR data and applied classifiers. Remote Sens-Basel 13:353. https://doi.org/10.3390/rs13030353
[37]	Mubin NA, Nadarajoo E, Shafri HZM, Hamedianfar A (2019) Young and mature oil palm tree detection and counting using convolutional neural network deep learning method. Int J Remote Sens 40:7500–7515. https://doi.org/10.1080/01431161.2019.1569282
[38]	Mullissa A, Vollrath A, Odongo-Braun C, Slagter B, Balling J, Gou YQ, Gorelick N, Reiche J (2021) Sentinel-1 SAR backscatter analysis ready data preparation in google earth engine. Remote Sens-Basel 13:1954. https://doi.org/10.3390/rs13101954
[39]	Mutanga O, Kumar L (2019) Google earth engine applications. Remote Sens-Basel 11:591. https://doi.org/10.3390/rs11050591
[40]	Nandasena WDK, Brabyn L, Serrao-Neumann S (2022) Using google earth engine to classify unique forest and agroforest classes using a mix of sentinel 2A spectral data and topographical features: a Sri Lanka case study. Geocarto Int 37:9544–9559. https://doi.org/10.1080/10106049.2021.2022010
[41]	Nicolau AP, Flores-Anderson A, Griffin R, Herndon K, Meyer FJ (2021) Assessing SAR C-band data to effectively distinguish modified land uses in a heavily disturbed Amazon forest. Int J Appl Earth Obs Geoinf 94:102214. https://doi.org/10.1016/j.jag.2020.102214
[42]	Pearman PB, Randin CF, Broennimann O, Vittoz P, Knaap WO, Engler R, Lay GL, Zimmermann NE, Guisan A (2008) Prediction of plant species distributions across six millennia. Ecol Lett 11:357–369. https://doi.org/10.1111/j.1461-0248.2007.01150.x
[43]	Pekel JF, Cottam A, Gorelick N, Belward AS (2016) High-resolution mapping of global surface water and its long-term changes. Nature 540:418–422. https://doi.org/10.1038/nature20584
[44]	Phiri D, Simwanda M, Salekin S, Nyirenda V, Murayama Y, Ranagalage M (2020) Sentinel-2 data for land cover/use mapping: a review. Remote Sens-Basel 12:2291. https://doi.org/10.3390/rs12142291
[45]	Pratico S, Solano F, Di Fazio S, Modica G (2021) Machine learning classification of Mediterranean forest habitats in google earth engine based on seasonal sentinel-2 time-series and input image composition optimisation. Remote Sens-Basel 13:586. https://doi.org/10.3390/rs13040586
[46]	Qin HM, Zhou WQ, Yao Y, Wang WM (2022) Individual tree segmentation and tree species classification in subtropical broadleaf forests using UAV-based LiDAR, hyperspectral, and ultrahigh-resolution RGB data. Remote Sens Environ 280:113143. https://doi.org/10.1016/j.rse.2022.113143
[47]	Rina S, Ying H, Shan Y, Du WL, Liu Y, Li R, Deng DZ (2023) Application of machine learning to tree species classification using active and passive remote sensing: a case study of the Duraer forestry zone. Remote Sens-Basel 15:2596. https://doi.org/10.3390/rs15102596
[48]	Schepaschenko D, See L, Lesiv M, Bastin JF, Mollicone D, Tsendbazar NE, Bastin L, McCallum L, Laso Bayas JC, Baklanov A, Perger C, Dürauer M, Fritz S (2019) Recent advances in forest observation with visual interpretation of very high-resolution imagery. Surv Geophys 40:839–862. https://doi.org/10.1007/s10712-019-09533-z
[49]	Schulze K, Malek Z, Verburg PH (2019) Towards better mapping of forest management patterns: a global allocation approach. For Ecol Manag 432:776–785. https://doi.org/10.1016/j.foreco.2018.10.001
[50]	Shao GF, Tang LN, Liao JF (2019) Overselling overall map accuracy misinforms about research reliability. Landsc Ecol 34:2487–2492. https://doi.org/10.1007/s10980-019-00916-6
[51]	Shao GF, Tang LN, Zhang H (2021) Introducing image classification efficacies. IEEE Access 9:134809–134816. https://doi.org/10.1109/ACCESS.2021.3116526
[52]	Shi YF, Wang TJ, Skidmore AK, Heurich M (2020) Improving LiDAR-based tree species mapping in central European mixed forests using multi-temporal digital aerial colour-infrared photographs. Int J Appl Earth Obs 84:101970. https://doi.org/10.1016/j.jag.2019.101970
[53]	Tamiminia H, Salehi B, Mahdianpari M, Quackenbush L, Adeli S, Brisco B (2020) Google earth engine for geo-big data applications: a meta-analysis and systematic review. ISPRS J Photogramm 164:152–170. https://doi.org/10.1016/j.isprsjprs.2020.04.001
[54]	Tan JW, Chang SW, Abdul-Kareem S, Yap HJ, Yong KT (2020) Deep learning for plant species classification using leaf vein morphometric. IEEE ACM T Comput Bi 17:82–90. https://doi.org/10.1109/TCBB.2018.2848653
[55]	Torres R, Snoeij P, Geudtner D, Bibby D, Davidson M, Attema E, Potin P, Rommen B, Floury N, Brown M, Traver IN, Deghaye P, Duesmann B, Rosich B, Miranda N, Bruno C, L’Abbate M, Croci R, Pietropaolo A, Huchler M, Rostan F (2012) GMES sentinel-1 mission. Remote Sens Environ 120:9–24. https://doi.org/10.1016/j.rse.2011.05.028
[56]	Vreugdenhil M, Wagner W, Bauer-Marschallinger B, Pfeil I, Teubner I, Rüdiger C, Strauss P (2018) Sensitivity of sentinel-1 backscatter to vegetation dynamics: an Austrian case study. Remote Sens-Basel 10:1396. https://doi.org/10.3390/rs10091396
[57]	Wang L, Diao CY, Xian G, Yin DM, Lu Y, Zou SY, Erickson TA (2020) A summary of the special issue on remote sensing of land change science with Google earth engine. Remote Sens Environ 248:112002. https://doi.org/10.1016/j.rse.2020.112002
[58]	Waser LT, Rüetschi M, Psomas A, Small D, Rehush N (2021) Mapping dominant leaf type based on combined sentinel-1/-2 data – Challenges for mountainous countries. ISPRS J Photogram 180:209–226. https://doi.org/10.1016/j.isprsjprs.2021.08.017
[59]	Xi ZX, Hopkinson C, Rood SB, Peddle DR (2020) See the forest and the trees: effective machine and deep learning algorithms for wood filtering and tree species classification from terrestrial laser scanning. ISPRS J Photogram 168:1–16. https://doi.org/10.1016/j.isprsjprs.2020.08.001
[60]	Xie B, Cao C, Xu M, Duerler RS, Yang XW, Bashir B, Chen YY, Wang KM (2021) Analysis of regional distribution of tree species using multi-seasonal sentinel-1 & 2 imagery within google earth engine. Forests 12:565. https://doi.org/10.3390/f12050565
[61]	Yao JX, Wu J, Xiao CZ, Zhang Z, Li JZ (2022) The classification method study of crops remote sensing with deep learning, machine learning, and google earth engine. Remote Sens-Basel 14:2758. https://doi.org/10.3390/rs14122758
[62]	Yuan QQ, Shen HF, Li TW, Li ZW, Li SW, Jiang Y, Xu HZ, Tan WW, Yang QQ, Wang JW, Gao JH, Zhang LP (2020) Deep learning in environmental remote sensing: achievements and challenges. Remote Sens Environ 241:111716. https://doi.org/10.1016/j.rse.2020.111716
[63]	Zanaga D, Van DKR, De KW, Souverijns N, Brockmann C, Quast R, Wevers J, Grosu A, Paccini A, Vergnaud S, Cartus O, Santoro M, Fritz S, Georgieva I, Lesiv M, Carter S, Herold M, Li LL, Tsendbazar NE, Ramoino F, Arino O (2021) ESA WorldCover 10 m 2020 v100 (Version v100). Zenodo. https://doi.org/10.5281/zenodo.5571936
[64]	Zhao Q, Yu L, Li XC, Peng DL, Zhang YG, Gong P (2021) Progress and trends in the application of google earth and google earth engine. Remote Sens-Basel 13:3778. https://doi.org/10.3390/rs13183778
[65]	Zheng PF, Fang PF, Wang LG, Ou GL, Xu WH, Dai F, Dai QL (2023) Synergism of multimodal data for mapping tree species distribution-a case study from a mountainous forest in southwest China. Remote Sens-Basel 15:979. https://doi.org/10.3390/rs15040979
[66]	Zhu Z, Qiu S, Ye S (2022) Remote sensing of land change: a multifaceted perspective. Remote Sens Environ 282:113266. https://doi.org/10.1016/j.rse.2022.113266
[67]	Zhu Z, Wang SX, Woodcock CE (2015) Improvement and expansion of the fmask algorithm: cloud, cloud shadow, and snow detection for landsats 4–7, 8, and sentinel 2 images. Remote Sens Environ 159:269–277. https://doi.org/10.1016/j.rse.2014.12.014