Mapping discrete forest age classes of Mediterranean pinelands since the pre-satellite era using historical orthoimage mosaics and machine learning

Abstract

Land use/land cover (LULC) change monitoring is critical for understanding environmental and socioeconomic processes and to identify patterns that may affect current and future land management. Forest cover evolution in the Mediterranean region has been studied to better understand forest succession, wildfires potential, and carbon stock assessment for climate change mitigation, among other reasons. However, though multiple sources of current LULC exist, data from last century’s forest cover are less common, and are normally still reliant on locally orthophoto-interpreted data, making continuous maps of historical forest cover relatively uncommon. In this work, a pipeline based on image segmentation and random forest LULC modeling was developed to process three high resolution orthophotos (1956, 1989, and 2021) into LULC continuous land cover maps of Spain’s island of Ibiza. Next, they were combined to quantify forest evolution of Mediterranean Aleppo pine (Pinus halepensis Mill.) and to generate a continuous map of forest age classes. Our models were able to differentiate forestland with an accuracy higher than 80% in all cases, and were able to approximate forestland cover change since the mid-twentieth century, estimating 21,165 ± 252 ha (37.0 ± 0.4%) in 1956, 27,099 ± 472 ha (46.8 ± 0.8%) in 1989, and 30,195 ± 302 ha (52.8 ± 0.5%) in 2021, with a mean increase of 139 ± 6 ha (0.46 ± 0.02%, calculated from current forest cover estimate) per year. The most important variables for the identification of the forestland were the terrain slope and the image gray level or color information in all orthophotos. When combining the information from the three periods, the analysis of forest evolution revealed that a significant portion of current forest cover, approximately 15,776 ha, fell within the 75–120 year age range, while 5388 ha fell within the range of 42–74 years, and 9022 ha within the 10–41 years forest age class. Younger forests, except when mapped after known wildfires, were not considered due to the limitations of the methodology. When compared to forest age data based on ground measurements, significant differences were found among each of the remotely sensed forest age classes, with a mean difference of 13 years between the theoretical age class central value and the real observed plot average age. Overall, 63% of the forest inventory plots were assigned with the correct forest age class. This work will allow a better understanding of long-term Mediterranean forest dynamics and will help landowners and policymakers to better respond to new landscape planning challenges and achieve sustainable development goals.

Keywords

Aerial orthophotos / Image segmentation / Random forest / Landscape evolution / Forest age

References

[1] AlberdiI, CañellasI, HernándezL, CondésS. A new method for the identification of old-growth trees in National Forest Inventories: application to Pinus halepensis Mill. stands in Spain. Ann for Sci, 2013, 70(3): 277-285

[2] Almonacid-CaballerJ, Pardo-PascualJE, RuizLA. Evaluating Fourier cross-correlation sub-pixel registration in landsat images. Remote Sens, 2017, 9(10): 1051

[3] AmbrosioL, AlonsoR. Estimación de superficies cultivadas por muestreo de áreas y teledetección. Precisión Relativa Estadística Española, 1993, 35(132): 91-103(in Spanish)

[4] AmezteguiA, Morán-OrdóñezA, MárquezA, Blázquez-CasadoÁ, PlaM, VilleroD, GarcíaMB, ErreaMP, CollL. Forest expansion in mountain protected areas: trends and consequences for the landscape. Landsc Urban Plan, 2021, 216: 104240

[5] AtikSO, IpbukerC. Integrating convolutional neural network and multiresolution segmentation for land cover and land use mapping using satellite imagery. Appl Sci, 2021, 11(12): 5551

[6] Bachtobji BouachirB, KhorchaniA, GuibalF, El AouniMH, KhaldiA. Dendroecological study of Pinus halepensis and Pinus pinea in northeast coastal dunes in Tunisia according to distance from the shoreline and dieback intensity. Dendrochronologia, 2017, 45: 62-72

[7] BajoccoS, De AngelisA, PeriniL, FerraraA, SalvatiL. The impact of land use/land cover changes on land degradation dynamics: a Mediterranean case study. Environ Manage, 2012, 49(5): 980-989

[8] BakkerMM, VeldkampA. Modelling land change: the issue of use and cover in wide-scale applications. J Land Use Sci, 2008, 3(4): 203-213

[9] BelwardAS, SkøienJO. Who launched what, when and why; trends in global land-cover observation capacity from civilian earth observation satellites. ISPRS J Photogramm Remote Sens, 2015, 103: 115-128

[10] BeyA, Sánchez-Paus DíazA, ManiatisD, MarchiG, MolliconeD, RicciS, BastinJF, MooreR, FedericiS, RezendeM, PatriarcaC, TuriaR, GamogaG, AbeH, KaidongE, MiceliG. Collect earth: land use and land cover assessment through augmented visual interpretation. Remote Sens, 2016, 8(10): 807

[11] BlaschkeT. Object based image analysis for remote sensing. ISPRS J Photogramm Remote Sens, 2010, 65(1): 2-16

[12] BossardM, FeranecJ, OtahelJCORINE land cover technical guide: addendum 2000, 2000, Copenhagen, European Environment Agency

[13] BreimanL, FriedmanJH, OlshenRA, StoneCJClassification and regression trees, 2017, Cambridge, Apple Academic Press

[14] BuddenbaumH, SchlerfM, HillJ. Classification of coniferous tree species and age classes using hyperspectral data and geostatistical methods. Int J Remote Sens, 2005, 26(24): 5453-5465

[15] BüttnerG, FeranecJ, JaffrainG, MariL, MauchaG, SoukupT. The CORINE land cover 2000 project. EARSeL eProceedings, 2004, 3: 331-346

[16] Camarero-RiojaLA. Tendencias recientes y evolución de la población rural en España. Política y Sociedad, 1991, 8: 13-24

[17] Castellnou M, Nebot E, Miralles M (2007) El papel del fuego en la gestión del paisaje. In: 4th International Wildland Fire Conference. Sevilla, pp. 13–17. (in Spanish)

[18] CastillaG, HayGJBlaschkeT, LangS, HayGJ. Image-objects and geographic objects. Object-Based Image Analysis, 2008, Heidelberg, Springer: 91-110

[19] ChenG, ThillJC, AnantsuksomsriS, TontisirinN, TaoR. Stand age estimation of rubber (Hevea brasiliensis) plantations using an integrated pixel- and object-based tree growth model and annual Landsat time series. ISPRS J Photogramm Remote Sens, 2018, 144: 94-104

[20] ChmielewskiS, BochniakA, NatapovA, WężykP. Introducing GEOBIA to landscape imageability assessment: a multi-temporal case study of the nature reserve “Kózki”. Poland Remote Sens, 2020, 12(17): 2792

[21] Cirer-Costa JC (2023) Ibiza and the introduction of capitalism in the Mediterranean: from agrarian self-consumption to full market dependence, 1857–1935. Historia Agraria 89: 159–191. https://doi.org/10.26882/histagrar.089e03c

[22] CMAT (2015) Conselleria d’Agricultura, Medi Ambient i Territori. IV Pla General de Defensa Contra Incendis Forestals de les Illes Balears. Govern de les Illes Balears, Tecnosylva SL.

[23] Collantes F (2007) La desagrarización de la sociedad rural española, 1950–1991. Historia Agraria – SEHA 42: 251–276.

[24] CostaH, FoodyGM, BoydDS. Supervised methods of image segmentation accuracy assessment in land cover mapping. Remote Sens Environ, 2018, 205: 338-351

[25] DasP, PandeyV. Use of logistic regression in land-cover classification with moderate-resolution multispectral data. J Indian Soc Remote Sens, 2019, 47(8): 1443-1454

[26] Delgado-ArtésR, Garófano-GómezV, Oliver-VillanuevaJV, Rojas-BrialesE. Land use/cover change analysis in the Mediterranean region: a regional case study of forest evolution in Castelló (Spain) over 50 years. Land Use Policy, 2022, 114: 105967

[27] Delibes-MateosM, FarfánMÁ, OliveroJ, VargasJM. Impact of land-use changes on red-legged partridge conservation in the Iberian Peninsula. Envir Conserv, 2012, 39(4): 337-346

[28] Di FilippoA, BalivaM, BrunettiM, Di FioreL. Long-term tree-ring response to drought and frost in two Pinus halepensis populations growing under contrasting environmental conditions in peninsular Italy. Forests, 2021, 12(3): 305

[29] DornikA, DrăguţL, UrdeaP. Classification of soil types using geographic object-based image analysis and random forests. Pedosphere, 2018, 28(6): 913-925

[30] Ez-zahouaniB, TeodoroA, El KharkiO, LiuJH, KotaridisI, YuanXH, MaL. Remote sensing imagery segmentation in object-based analysis: a review of methods, optimization, and quality evaluation over the past 20 years. Remote Sens Appl Soc Environ, 2023, 32: 101031

[31] Di Fazio S, Modica G, Zoccali P (2011) Evolution trends of land use/land cover in a mediterranean forest landscape in Italy. In: Murgante B, Gervasi O, Iglesias A, Taniar D, Apduhan BO (Eds.) Comp. Sci. Applic. - ICCSA 2011. ICCSA 2011. Lecture Notes in Computer Science 6782. Springer, Berlin, Heidelberg. https://doi.org/10.1007/978-3-642-21928-3_20

[32] Fernández GarcíaF. Fotografía aérea histórica e historia de la fotografía aérea en España. Ería, 2015, 98(98): 217-240

[33] FoxDM, WitzE, BlancV, SouliéC, Penalver-NavarroM, DervieuxA. A case study of land cover change (1950–2003) and runoff in a Mediterranean catchment. Appl Geogr, 2012, 32(2): 810-821

[34] FrateL, CarranzaML, GarfìV, FebbraroMD, TontiD, MarchettiM, OttavianoM, SantopuoliG, ChiriciG. Spatially explicit estimation of forest age by integrating remotely sensed data and inverse yield modeling techniques. Iforest, 2016, 9(1): 63-71

[35] FraverS, BradfordJB, PalikBJ. Improving tree age estimates derived from increment cores: a case study of red pine. For Sci, 2011, 57(2): 164-170

[36] FuZ, LiDJ, HararukO, SchwalmC, LuoYQ, YanLM, NiuSL. Recovery time and state change of terrestrial carbon cycle after disturbance. Environ Res Lett, 2017, 12(10): 104004

[37] García-RuizJM, RegüésD, AlveraB, Lana-RenaultN, Serrano-MuelaP, Nadal-RomeroE, NavasA, LatronJ, Martí-BonoC, ArnáezJ. Flood generation and sediment transport in experimental catchments affected by land use changes in the central Pyrenees. J Hydrol, 2008, 356(1–2): 245-260

[38] GómezC, WhiteJC, WulderMA. Optical remotely sensed time series data for land cover classification: a review. ISPRS J Photogramm Remote Sens, 2016, 116: 55-72

[39] González-ÁvilaS, López-LeivaC, BunceRGH, Elena-RossellóR. Changes and drivers in Spanish landscapes at the rural-urban interface between 1956 and 2018. Sci Total Environ, 2020, 714: 136858

[40] González-Avila S, Elena-Rosselló R (2009) SISPARES: cartografía nacional del patrón y cambio de los paisajes rurales españoles. Cuadernos de la Sociedad Española de Ciencias Forestales 29: 15–21. https://doi.org/10.31167/csef.v0i29.9812

[41] GromadaK. Unsupervised SAR imagery feature learning with Median filter-based loss value. Sensors, 2022, 22(17): 6519

[42] GrottoliE, BiausqueM, RogersD, JacksonDWT, CooperJAG. Structure-from-motion-derived digital surface models from historical aerial photographs: a new 3D application for coastal dune monitoring. Remote Sens, 2021, 13(1): 95

[43] GuzhaAC, RufinoMC, OkothS, JacobsS, NóbregaRLB. Impacts of land use and land cover change on surface runoff, discharge and low flows: evidence from East Africa. J Hydrol Reg Stud, 2018, 15: 49-67

[44] HeisigH, SimmenJL. Re-engineering the past: countrywide geo-referencing of archival aerial imagery. PFG, 2021, 89(6): 487-503

[45] Joint Research Centre (2020) Global map of forest cover for year 2020. EU observatory on deforestation and forest degradation. European Commission.

[46] KaimD, KozakJ, KoleckaN, ZiółkowskaE, OstafinK, OstapowiczK, GimmiU, MunteanuC, RadeloffVC. Broad scale forest cover reconstruction from historical topographic maps. Appl Geogr, 2016, 67: 39-48

[47] KayitakireF, HamelC, DefournyP. Retrieving forest structure variables based on image texture analysis and IKONOS-2 imagery. Remote Sens Environ, 2006, 102(3–4): 390-401

[48] KennedyRE, YangZQ, CohenWB. Detecting trends in forest disturbance and recovery using yearly Landsat time series: 1 LandTrendr—Temporal segmentation algorithms. Remote Sens Environ, 2010, 114(12): 2897-2910

[49] KnuthF, SheanD, BhushanS, SchwatE, AlexandrovO, McNeilC, DehecqA, FlorentineC, O’NeelS. Historical structure from motion (HSfM): automated processing of historical aerial photographs for long-term topographic change analysis. Remote Sens Environ, 2023, 285: 113379

[50] KochAJ, DriscollDA, KirkpatrickJB. Estimating the accuracy of tree ageing methods in mature Eucalyptus obliqua forest. Tasmania Aust for, 2008, 71(2): 147-159

[51] KwonSK, JungHS, BaekWK, KimD. Classification of forest vertical structure in south Korea from aerial orthophoto and lidar data using an artificial neural network. Appl Sci, 2017, 7(10): 1046

[52] LangenkampJP, RienowA. Exploring the use of orthophotos in google earth engine for very high-resolution mapping of impervious surfaces: a data fusion approach in Wuppertal. Germany Remote Sens, 2023, 15(7): 1818

[53] Lasanta-MartínezT, Vicente-SerranoSM, Cuadrat-PratsJM. Mountain Mediterranean landscape evolution caused by the abandonment of traditional primary activities: a study of the Spanish Central Pyrenees. Appl Geogr, 2005, 25(1): 47-65

[54] LiuDS, XiaF. Assessing object-based classification: advantages and limitations. Remote Sens Lett, 2010, 1(4): 187-194

[55] LiuXP, HeJL, YaoY, ZhangJB, LiangHL, WangH, HongY. Classifying urban land use by integrating remote sensing and social media data. Int J Geogr Inf Sci, 2017, 31(8): 1675-1696

[56] LiuK, ShenX, CaoL, WangGB, CaoFL. Estimating forest structural attributes using UAV-LiDAR data in Ginkgo plantations. ISPRS J Photogramm Remote Sens, 2018, 146: 465-482

[57] LuH, LiuC, LiNW, FuX, LiLG. Optimal segmentation scale selection and evaluation of cultivated land objects based on high-resolution remote sensing images with spectral and texture features. Environ Sci Pollut Res Int, 2021, 28(21): 27067-27083

[58] MAGRAMA (2012) Ministerio de Agricultura, Alimentación y Medio Ambiente. 4th Spanish national forest inventory. Dirección General de Desarrollo Rural y Política Forestal, Área de Inventario y Estadísticas Forestales.

[59] MaltamoM, KinnunenH, KangasA, KorhonenL. Predicting stand age in managed forests using National Forest Inventory field data and airborne laser scanning. For Ecosyst, 2020, 7: 44

[60] MaltmanJC, HermosillaT, WulderMA, CoopsNC, WhiteJC. Estimating and mapping forest age across Canada’s forested ecosystems. Remote Sens Environ, 2023, 290: 113529

[61] Martínez-FernándezV, del TánagoMG, MarotoJ, de JalónDG. Fluvial corridor changes over time in regulated and non-regulated rivers (upper esla river, NW Spain). River Res Appl, 2017, 33(2): 214-223

[62] MaxwellAE, StragerMP, WarnerTA, RamezanCA, MorganAN, PauleyCE. Large-area, high spatial resolution land cover mapping using random forests, GEOBIA, and NAIP orthophotography: findings and recommendations. Remote Sens, 2019, 11(12): 1409

[63] MetsarantaJM. Dendrochronological procedures improve the precision and accuracy of tree and stand age estimates in the Western Canadian boreal forest. For Ecol Manag, 2020, 457: 117657

[64] MolnarC, BischlB, CasalicchioG. iml: An R package for interpretable machine learning. J Open Source Soft, 2018, 3(26): 786

[65] More AS, Rana DP (2017) Review of random forest classification techniques to resolve data imbalance. In: 2017 1st International Conference on Intelligent Systems and Information Management (ICISIM). Aurangabad, India. IEEE 72–78. https://doi.org/10.1109/ICISIM.2017.8122151

[66] MorfordSL, AllredBW, JensenER, MaestasJD, MuellerKR, PacholskiCL, SmithJT, TackJD, TackettKN, NaugleDE. Mapping tree cover expansion in Montana, USA rangelands using high-resolution historical aerial imagery. Remote Sens Ecol Conserv, 2024, 10(1): 91-105

[67] NguyenTA, KellenbergerB, TuiaD. Mapping forest in the Swiss Alps treeline ecotone with explainable deep learning. Remote Sens Environ, 2022, 281: 113217

[68] NguyenTA, RußwurmM, LencznerG, TuiaD. Multi-temporal forest monitoring in the Swiss Alps with knowledge-guided deep learning. Remote Sens Environ, 2024, 305: 114109

[69] NitaMD, MunteanuC, GutmanG, AbrudanIV, RadeloffVC. Widespread forest cutting in the aftermath of World War II captured by broad-scale historical Corona spy satellite photography. Remote Sens Environ, 2018, 204: 322-332

[70] Oliver CD, Larson CB (1996) Forest stand dynamics: updated edition. Environmental Science. John Wiley and sons.

[71] PatinoJE, DuqueJC. A review of regional science applications of satellite remote sensing in urban settings. Comput Environ Urban Syst, 2013, 37: 1-17

[72] PintoAT, GonçalvesJA, BejaP, Pradinho HonradoJ. From archived historical aerial imagery to informative orthophotos: a framework for retrieving the past in long-term socioecological research. Remote Sens, 2019, 11(11): 1388

[73] PonsA, RullanO. Artificialization and islandness on the Spanish tourist coast. Misc Geogr Reg Stud Dev, 2014, 18(1): 5-16

[74] PonsA, RullanO. The expansion of urbanisation in the balearic islands (1956–2006). J Mar Isl Cult, 2014, 3(2): 78-88

[75] ProbstP, WrightMN, BoulesteixAL. Hyperparameters and tuning strategies for random forest. Wiley Interdiscip Rev Data Min Knowl Discov, 2019, 9(3): e1301

[76] QGIS Development Team (2022) QGIS geographic information system. QGIS Association. https://www.qgis.org

[77] Ramón-Cardona J, Serra-Cantallops A (2014) Historia del turismo en Ibiza: aplicación del ciclo de Vida del destino turístico en un destino maduro del mediterráneo. PASOS Rev De Turismo Y Patrimonio Cult 12(4): 899–913. https://doi.org/10.25145/j.pasos.2014.12.06

[78] Reyes-PalomequeG, DupuyJM, Portillo-QuinteroCA, AndradeJL, Tun-DzulFJ, Hernández-StefanoniJL. Mapping forest age and characterizing vegetation structure and species composition in tropical dry forests. Ecol Indic, 2021, 120: 106955

[79] Ribas-CostaVA, CookRL, GastónA. Modeling structural traits of Aleppo pine (Pinus halepensis Mill) forests with low-density LiDAR. Eur J Remote Sens, 2024, 57(1): 2344569

[80] Del Río M, Calama R, Montero G (2008) Selvicultura de Pinus halepensis Mill. in “Compendio de Selvicultura Aplicada en España”, Serrada R, Montero G, Reque JA (Eds.) Madrid: Instituto Nacional de investigación y Tecnología Agraria y Alimentaria. Madrid, Spain.

[81] Roig-Coll S, Kechaiga A (2022) A land of extremes: the dance between intensive farming and farmland abandonment. In: “Resources, Strategis and Potentialities for Food System Sustainability in the Mediterranean Area. Bertoret N, Bertoret E (Eds). Ajuntament de València, CEMAS. València, Spain.

[82] Rojas-Briales E. Delgado-Artés R, Cabrera-Bonet M (2018) Human desertification and disempowerment of rural territories. Forestry in the Midst of Global Changes. CRC Press 101–134

[83] RóżyckiS, KarwelAK, KurczyńskiZ. German extermination camps on WWII reconnaissance photographs. Orthorectification process for archival aerial images of cultural heritage sites. Remote Sens, 2023, 15(10): 2587

[84] SchumacherJ, HauglinM, AstrupR, BreidenbachJ. Mapping forest age using National Forest Inventory, airborne laser scanning, and Sentinel-2 data. For Ecosyst, 2020, 7: 60

[85] SharmaA, LiuXW, YangXJ. Land cover classification from multi-temporal, multi-spectral remotely sensed imagery using patch-based recurrent neural networks. Neural Netw, 2018, 105: 346-355

[86] ShimizuK, MurakamiW, FuruichiT, EstoqueRC. Mapping land use/land cover changes and forest disturbances in Vietnam using a landsat temporal segmentation algorithm. Remote Sens, 2023, 15(3): 851

[87] SmithA. Image segmentation scale parameter optimization and land cover classification using the Random Forest algorithm. J Spatial Sci, 2010, 55(1): 69-79

[88] SohnJA, SahaS, BauhusJ. Potential of forest thinning to mitigate drought stress: a meta-analysis. For Ecol Manag, 2016, 380: 261-273

[89] SouzaCMJr, ShimboJZ, RosaMR, ParenteLL, AlencarAA, RudorffBFT, HasenackH, MatsumotoM, FerreiraLG, Souza-FilhoPWM, de OliveiraSW, RochaWF, FonsecaAV, MarquesCB, DinizCG, CostaD, MonteiroD, RosaER, Vélez-MartinE, WeberEJ, LentiFEB, PaternostFF, PareynFGC, SiqueiraJV, VieraJL, NetoLCF, SaraivaMM, SalesMH, SalgadoMPG, VasconcelosR, GalanoS, MesquitaVV, AzevedoT. Reconstructing three decades of land use and land cover changes in Brazilian biomes with landsat archive and earth engine. Remote Sens, 2020, 12(17): 2735

[90] Speer JH (2010) Fundamentals of tree-ring research. University of Arizona Press 368. https://doi.org/10.5860/choice.48-1462

[91] StehmanSV. Estimating area and map accuracy for stratified random sampling when the strata are different from the map classes. Int J Remote Sens, 2014, 35(13): 4923-4939

[92] Stevenson M, Sergeant E, Firestone S (2023) epiR: tools for the analysis of epidemiological data. R package version 2.0.66. https://CRAN.R-project.org/package=epiR

[93] TadeseM, KumarL, KoechR, KogoBK. Mapping of land-use/land-cover changes and its dynamics in Awash River Basin using remote sensing and GIS. Remote Sens Appl Soc Environ, 2020, 19: 100352

[94] TassiA, VizzariM. Object-oriented LULC classification in google earth engine combining SNIC, GLCM, and machine learning algorithms. Remote Sens, 2020, 12(22): 3776

[95] ThomasVA, WynneRH, KauffmanJ, McCurdyW, BrooksEB, ThomasRQ, RakestrawJ. Mapping thins to identify active forest management in southern pine plantations using Landsat time series stacks. Remote Sens Environ, 2021, 252: 112127

[96] TolanJ, YangHI, NosarzewskiB, CouaironG, VoHV, BrandtJ, SporeJ, MajumdarS, HazizaD, VamarajuJ, MoutakanniT, BojanowskiP, JohnsT, WhiteB, TieckeT, CouprieC. Very high resolution canopy height maps from RGB imagery using self-supervised vision transformer and convolutional decoder trained on aerial lidar. Remote Sens Environ, 2024, 300: 113888

[97] Van den BroeckWAJ, GoedeméT, LoopmansM. Multiclass land cover mapping from historical orthophotos using domain adaptation and spatio-temporal transfer learning. Remote Sens, 2022, 14(23): 5911

[98] Varo-MartínezMÁ, Navarro-CerrilloRM, Hernández-ClementeR, Duque-LazoJ. Semi-automated stand delineation in Mediterranean Pinus sylvestris plantations through segmentation of LiDAR data: the influence of pulse density. Int J Appl Earth Obs Geoinf, 2017, 56: 54-64

[99] VastarantaM, NiemiM, WulderMA, WhiteJC, NurminenK, LitkeyP, HonkavaaraE, HolopainenM, HyyppäJ. Forest stand age classification using time series of photogrammetrically derived digital surface models. Scand J Res, 2016, 31(2): 194-205

[100] VégaC, St-OngeB. Mapping site index and age by linking a time series of canopy height models with growth curves. For Ecol Manag, 2009, 257(3): 951-959

[101] Vidal-MacuaJJ, NinyerolaM, ZabalaA, Domingo-MarimonC, Gonzalez-GuerreroO, PonsX. Environmental and socioeconomic factors of abandonment of rainfed and irrigated crops in northeast Spain. Appl Geogr, 2018, 90: 155-174

[102] WangYJ, QiQW, LiuY, JiangLL, WangJ. Unsupervised segmentation parameter selection using the local spatial statistics for remote sensing image segmentation. Int J Appl Earth Obs Geoinf, 2019, 81: 98-109

[103] Weiss A (2001) Topographic position and landforms analysis. Poster presentation, ESRI User Conference, San Diego, CA

[104] WężykP, HawryloP, JanusB, WeidenbachM, SzostakM. Forest cover changes in Gorce NP (Poland) using photointerpretation of analogue photographs and GEOBIA of orthophotos and nDSM based on image-matching based approach. Eur J Remote Sens, 2018, 51(1): 501-510

[105] WrightMN, ZieglerA. ranger: a fast implementation of random forests for high dimensional data in C++ and R. J Stat Soft, 2017, 77(1): 1-17

[106] WulderMA, FranklinSE, WhiteJC, LinkeJL, MagnussenS. An accuracy assessment framework for large-area land cover classification products derived from medium–—resolution satellite data. Int J Remote Sens, 2006, 27(4): 663-683

[107] WulderMA, WhiteJC, LovelandTR, WoodcockCE, BelwardAS, CohenWB, FosnightEA, ShawJ, MasekJG, RoyDP. The global Landsat archive: status, consolidation, and direction. Remote Sens Environ, 2016, 185: 271-283

[108] YangXG, LiuYQ, WuZC, YuY, LiFR, FanWY. Forest age mapping based on multiple-resource remote sensing data. Environ Monit Assess, 2020, 192(11): 734

[109] ZhangC, SargentI, PanX, LiHP, GardinerA, HareJ, AtkinsonPM. Joint deep learning for land cover and land use classification. Remote Sens Environ, 2019, 221: 173-187

[110] ZhouHY, FuLY, SharmaRP, LeiYC, GuoJP. A hybrid approach of combining random forest with texture analysis and VDVI for desert vegetation mapping based on UAV RGB data. Remote Sens, 2021, 13(10): 1891

Funding

Universidad Politécnica de Madrid

RIGHTS & PERMISSIONS

The Author(s)