|
|
Registro completo
|
Biblioteca (s) : |
INIA Las Brujas. |
Fecha : |
16/03/2022 |
Actualizado : |
16/03/2022 |
Tipo de producción científica : |
Artículos en Revistas Indexadas Internacionales |
Autor : |
HIRIGOYEN, A.; ACOSTA-MUÑOZ, C.; SALAMANCA, A.J.A.; VARO-MARTINEZ, M.Á.; RACHID, C.; FRANCO, J.; NAVARRO-CERRILLO, R. |
Afiliación : |
ANDRES EDUARDO HIRIGOYEN DOMINGUEZ, INIA (Instituto Nacional de Investigación Agropecuaria), Uruguay; CRISTINA ACOSTA-MUÑOZ, Department of Forestry Engineering, Laboratory of Silviculture, Dendrochronology and Climate Change, DendrodatLab-ERSAF, University of Cordoba, Córdoba, Spain; ANTONIO JESÚS ARIZA SALAMANCA, Department of Forestry Engineering, Laboratory of Silviculture, Dendrochronology and Climate Change, DendrodatLab-ERSAF, University of Cordoba, Córdoba, Spain; MARIA ÁNGELES VARO-MARTINEZ, Department of Forestry Engineering, Laboratory of Silviculture, Dendrochronology and Climate Change, DendrodatLab-ERSAF, University of Cordoba, Córdoba, Spain; ANA CECILIA RACHID CASNATI, INIA (Instituto Nacional de Investigación Agropecuaria), Uruguay; JORGE FRANCO, Faculty of Agronomy, University of the Republic, Paysandú, Uruguay; RAFAEL NAVARRO-CERRILLO, Department of Forestry Engineering, Laboratory of Silviculture, Dendrochronology and Climate Change, DendrodatLab-ERSAF, University of Cordoba, Córdoba, Spain. |
Título : |
A machine learning approach to model leaf area index in Eucalyptus plantations using high-resolution satellite imagery and airborne laser scanner data. |
Fecha de publicación : |
2021 |
Fuente / Imprenta : |
Annals of Forest Research, 2021, Volume 64, Issue 2, Pages 165-183. OPEN ACCESS. doi: https://doi.org/10.15287/afr.2021.2073 |
ISSN : |
1844-8135 |
DOI : |
10.15287/afr.2021.2073 |
Idioma : |
Inglés |
Notas : |
Article history: Received October 27, 2020; Revised December 14, 2021; Accepted December 21, 2021.
Corresponding author: Hirigoyen, A.; National Institute of Agricultural Research (INIA), Tacuarembó, Uruguay; email:ahirigoyen@inia.org.uy -- The authors thank the Instituto Nacional de Investigaciones Agropecuarias (INIA-Uruguay) for supporting our research work and for help during the fieldwork. We are particularly grateful to Roberto Scoz, Demian Gomez, Leonidas Carrasco and Alicia Peduzzi for their assistance during this research. RMNC acknowledge the institutional support of the Ministerio de Ciencia, Innovaci?n y Universidades (Spain), through the ESPECTRAMED (CGL2017-86161-R). show significant changes. |
Contenido : |
ABSTRACT. - As a forest structural parameter, leaf area index (LAI) is crucial for efficient intensive plantation management. Leaf area is responsible for the energy absorption needed for photosynthetic production and transpiration, both affecting growth. Currently, LAI can be estimated either by remote-sensing methods or ground-based methods. However, unlike ground-based methods, remote estimation provides a cost-effective and ecologically significant advance. The aim of our study was to evaluate whether machine learning algorithms can be used to quantify LAI, using either optical remote sensing or LiDAR metrics in Eucalyptus dunnii and Eucalyptus grandis stands. First, empirical relationships between LAI and remote-sensing data using LiDAR metrics and multispectral high-resolution satellite metrics, were assessed. Selected variables for LAI estimation were: forest canopy cover, laser penetration index, canopy relief ratio (from among the LiDAR data), the green normalized difference vegetation index, and normalized difference vegetation index (from among spectral vegetation indices). We compared the accuracy of three machine learning algorithms: artificial neural networks (ANN), random forest (RF) and support vector regression (SVR). The coefficient of determination ranged from 0.60, for ANN, to 0.84, for SVR. The SVR regression methods showed the best performance in terms of overall model accuracy and RMSE (0.60). The results show that the remote sensing data applied through machine learning algorithms provide an effective method to estimate LAI in eucalypt plantations. The methodology proposed is directly applicable for operational forest planning at the landscape level. © 2021, Editura Silvica. All rights reserved. MenosABSTRACT. - As a forest structural parameter, leaf area index (LAI) is crucial for efficient intensive plantation management. Leaf area is responsible for the energy absorption needed for photosynthetic production and transpiration, both affecting growth. Currently, LAI can be estimated either by remote-sensing methods or ground-based methods. However, unlike ground-based methods, remote estimation provides a cost-effective and ecologically significant advance. The aim of our study was to evaluate whether machine learning algorithms can be used to quantify LAI, using either optical remote sensing or LiDAR metrics in Eucalyptus dunnii and Eucalyptus grandis stands. First, empirical relationships between LAI and remote-sensing data using LiDAR metrics and multispectral high-resolution satellite metrics, were assessed. Selected variables for LAI estimation were: forest canopy cover, laser penetration index, canopy relief ratio (from among the LiDAR data), the green normalized difference vegetation index, and normalized difference vegetation index (from among spectral vegetation indices). We compared the accuracy of three machine learning algorithms: artificial neural networks (ANN), random forest (RF) and support vector regression (SVR). The coefficient of determination ranged from 0.60, for ANN, to 0.84, for SVR. The SVR regression methods showed the best performance in terms of overall model accuracy and RMSE (0.60). The results show that the remote sensing data applied throu... Presentar Todo |
Palabras claves : |
Intensive silviculture; LAI canopy; Machine learning algorithms. |
Asunto categoría : |
K01 Ciencias forestales - Aspectos generales |
URL : |
https://www.afrjournal.org/index.php/afr/article/viewFile/2073/1177
|
Marc : |
LEADER 03380naa a2200265 a 4500 001 1062842 005 2022-03-16 008 2021 bl uuuu u00u1 u #d 022 $a1844-8135 024 7 $a10.15287/afr.2021.2073$2DOI 100 1 $aHIRIGOYEN, A. 245 $aA machine learning approach to model leaf area index in Eucalyptus plantations using high-resolution satellite imagery and airborne laser scanner data.$h[electronic resource] 260 $c2021 500 $aArticle history: Received October 27, 2020; Revised December 14, 2021; Accepted December 21, 2021. Corresponding author: Hirigoyen, A.; National Institute of Agricultural Research (INIA), Tacuarembó, Uruguay; email:ahirigoyen@inia.org.uy -- The authors thank the Instituto Nacional de Investigaciones Agropecuarias (INIA-Uruguay) for supporting our research work and for help during the fieldwork. We are particularly grateful to Roberto Scoz, Demian Gomez, Leonidas Carrasco and Alicia Peduzzi for their assistance during this research. RMNC acknowledge the institutional support of the Ministerio de Ciencia, Innovaci?n y Universidades (Spain), through the ESPECTRAMED (CGL2017-86161-R). show significant changes. 520 $aABSTRACT. - As a forest structural parameter, leaf area index (LAI) is crucial for efficient intensive plantation management. Leaf area is responsible for the energy absorption needed for photosynthetic production and transpiration, both affecting growth. Currently, LAI can be estimated either by remote-sensing methods or ground-based methods. However, unlike ground-based methods, remote estimation provides a cost-effective and ecologically significant advance. The aim of our study was to evaluate whether machine learning algorithms can be used to quantify LAI, using either optical remote sensing or LiDAR metrics in Eucalyptus dunnii and Eucalyptus grandis stands. First, empirical relationships between LAI and remote-sensing data using LiDAR metrics and multispectral high-resolution satellite metrics, were assessed. Selected variables for LAI estimation were: forest canopy cover, laser penetration index, canopy relief ratio (from among the LiDAR data), the green normalized difference vegetation index, and normalized difference vegetation index (from among spectral vegetation indices). We compared the accuracy of three machine learning algorithms: artificial neural networks (ANN), random forest (RF) and support vector regression (SVR). The coefficient of determination ranged from 0.60, for ANN, to 0.84, for SVR. The SVR regression methods showed the best performance in terms of overall model accuracy and RMSE (0.60). The results show that the remote sensing data applied through machine learning algorithms provide an effective method to estimate LAI in eucalypt plantations. The methodology proposed is directly applicable for operational forest planning at the landscape level. © 2021, Editura Silvica. All rights reserved. 653 $aIntensive silviculture 653 $aLAI canopy 653 $aMachine learning algorithms 700 1 $aACOSTA-MUÑOZ, C. 700 1 $aSALAMANCA, A.J.A. 700 1 $aVARO-MARTINEZ, M.Á. 700 1 $aRACHID, C. 700 1 $aFRANCO, J. 700 1 $aNAVARRO-CERRILLO, R. 773 $tAnnals of Forest Research, 2021, Volume 64, Issue 2, Pages 165-183. OPEN ACCESS. doi: https://doi.org/10.15287/afr.2021.2073
Descargar
Esconder MarcPresentar Marc Completo |
Registro original : |
INIA Las Brujas (LB) |
|
Biblioteca
|
Identificación
|
Origen
|
Tipo / Formato
|
Clasificación
|
Cutter
|
Registro
|
Volumen
|
Estado
|
Volver
|
|
| Acceso al texto completo restringido a Biblioteca INIA Las Brujas. Por información adicional contacte bibliolb@inia.org.uy. |
Registro completo
|
Biblioteca (s) : |
INIA Las Brujas; INIA Treinta y Tres. |
Fecha actual : |
12/11/2015 |
Actualizado : |
09/10/2019 |
Tipo de producción científica : |
Artículos en Revistas Indexadas Internacionales |
Circulación / Nivel : |
Internacional - -- |
Autor : |
MARCAIDA, M.; ASSENG, S.; EWERT, F.; BASSU, S.; DURAND, J.L.; LI, T.; MARTRE, P.; ADAM, M.; AGGARWAL, P.K.; ANGULO, C.; BARON, C.; BASSO, B.; BERTUZZI, P.; BIERNATH, C.; BOOGAARD, H.; BOOTE, K.J.; BOUMAN, B.; BREGAGLIO, S.; BRISSON, N.; BUIS, S.; CAMMARANO, D.; CHALLINOR, A.J.; CONFALONIERI, R.; CONIJN, J.G.; CORBEELS, M.; DERYNG, D.; DE SANCTIS, G.; DOLTRA, J.; FUMOTO, T.; GAYDON, D.; GAYLER, S.; GOLDBERG, R.; GRANT, R.F.; GRASSINI, P.; HATFIELD, J.L.; HASEGAWA, T.; HENG, L.; HOEK, S.; HOOKER, J.; HUNT, L.A.; INGWERSEN, J.; IZAURRALDE, R.C.; JONGSCHAAP, R.E.E.; JONES, J.W.; KEMANIAN, R.A.; KERSEBAUM, K.C.; KIM, S.-H.; LIZASO, J.; MÜLLER, C.; NAKAGAWA, H.; NARESH KUMAR, S.; NENDEL, C.; O'LEARY, G.J.; OLESEN, J.E.; ORIOL, P.; OSBORNE, T.M.; PALOSUO, T.; PRAVIA, V.; PRIESACK, E.; RIPOCHE, D.; ROSENZWEIG, C.; RUANE, A.C.; RUGET, F.; SAU, F.; SEMENOV, M.A.; SHCHERBAK, I.; SINGH, B.; SINGH, U.; SOO, H.K.; STEDUTO, P.; STÖCKLE, C.; STRATONOVITCH, P.; STRECK, T.; SUPIT, I.; TANG, L.; TAO, F.; TEIXEIRA, E.I.; THORBURN, P.; TIMLIN, D.; TRAVASSO, M.; RÖTTER, R.P.; WAHA, K.; WALLACH, D.; WHITE, J.W.; WILKENS, P.; WILLIAMS, J.R.; WOLF, J.; YIN, X.; YOSHIDA, H.; ZHANG, Z.; ZHU, Y. |
Afiliación : |
MARIA VIRGINIA PRAVIA NIN, INIA (Instituto Nacional de Investigación Agropecuaria), Uruguay. |
Título : |
A statistical analysis of three ensembles of crop model responses to temperature and CO2 concentration. |
Fecha de publicación : |
2015 |
Fuente / Imprenta : |
Agricultural and Forest Meteorology, 2015, v.214-215, p. 483-493. |
ISSN : |
0168-1923 |
DOI : |
10.1016/j.agrformet.2015.09.013 |
Idioma : |
Inglés |
Notas : |
Article history: Received 6 March 2015 / Received in revised form 29 July 2015 / Accepted 20 September 2015 / Available online 1 October 2015. |
Contenido : |
ABSTRACT.
Ensembles of process-based crop models are increasingly used to simulate crop growth for scenariosof temperature and/or precipitation changes corresponding to different projections of atmospheric CO2concentrations. This approach generates large datasets with thousands of simulated crop yield data. Suchdatasets potentially provide new information but it is difficult to summarize them in a useful way due totheir structural complexities. An associated issue is that it is not straightforward to compare crops and tointerpolate the results to alternative climate scenarios not initially included in the simulation protocols.Here we demonstrate that statistical models based on random-coefficient regressions are able to emulateensembles of process-based crop models. An important advantage of the proposed statistical models isthat they can interpolate between temperature levels and between CO2concentration levels, and canthus be used to calculate temperature and [CO2] thresholds leading to yield loss or yield gain, without re-running the original complex crop models. Our approach is illustrated with three yield datasets simulatedby 19 maize models, 26 wheat models, and 13 rice models. Several statistical models are fitted to thesedatasets, and are then used to analyze the variability of the yield response to [CO2] and temperature.Based on our results, we show that, for wheat, a [CO2] increase is likely to outweigh the negative effectof a temperature increase of +2◦C in the considered sites. Compared to wheat, required levels of [CO2]increase are much higher for maize, and intermediate for rice. For all crops, uncertainties in simulatingclimate change impacts increase more with temperature than with elevated [CO2].
© 2015 Elsevier B.V. All rights reserved. MenosABSTRACT.
Ensembles of process-based crop models are increasingly used to simulate crop growth for scenariosof temperature and/or precipitation changes corresponding to different projections of atmospheric CO2concentrations. This approach generates large datasets with thousands of simulated crop yield data. Suchdatasets potentially provide new information but it is difficult to summarize them in a useful way due totheir structural complexities. An associated issue is that it is not straightforward to compare crops and tointerpolate the results to alternative climate scenarios not initially included in the simulation protocols.Here we demonstrate that statistical models based on random-coefficient regressions are able to emulateensembles of process-based crop models. An important advantage of the proposed statistical models isthat they can interpolate between temperature levels and between CO2concentration levels, and canthus be used to calculate temperature and [CO2] thresholds leading to yield loss or yield gain, without re-running the original complex crop models. Our approach is illustrated with three yield datasets simulatedby 19 maize models, 26 wheat models, and 13 rice models. Several statistical models are fitted to thesedatasets, and are then used to analyze the variability of the yield response to [CO2] and temperature.Based on our results, we show that, for wheat, a [CO2] increase is likely to outweigh the negative effectof a temperature increase of +2◦C in... Presentar Todo |
Palabras claves : |
Climate change; CROP MODEL; Emulator; MAIZE; Meta-model; MODELIZACIÓN DE LOS CULTIVOS; RICE; Statistical model; WHEAT; Yield. |
Thesagro : |
ARROZ; CAMBIO CLIMÁTICO; MAÍZ; MODELOS ESTADISTICOS; TRIGO. |
Asunto categoría : |
A50 Investigación agraria |
Marc : |
LEADER 05363naa a2201417 a 4500 001 1053856 005 2019-10-09 008 2015 bl uuuu u00u1 u #d 022 $a0168-1923 024 7 $a10.1016/j.agrformet.2015.09.013$2DOI 100 1 $aMARCAIDA, M. 245 $aA statistical analysis of three ensembles of crop model responses to temperature and CO2 concentration. 260 $c2015 500 $aArticle history: Received 6 March 2015 / Received in revised form 29 July 2015 / Accepted 20 September 2015 / Available online 1 October 2015. 520 $aABSTRACT. Ensembles of process-based crop models are increasingly used to simulate crop growth for scenariosof temperature and/or precipitation changes corresponding to different projections of atmospheric CO2concentrations. This approach generates large datasets with thousands of simulated crop yield data. Suchdatasets potentially provide new information but it is difficult to summarize them in a useful way due totheir structural complexities. An associated issue is that it is not straightforward to compare crops and tointerpolate the results to alternative climate scenarios not initially included in the simulation protocols.Here we demonstrate that statistical models based on random-coefficient regressions are able to emulateensembles of process-based crop models. An important advantage of the proposed statistical models isthat they can interpolate between temperature levels and between CO2concentration levels, and canthus be used to calculate temperature and [CO2] thresholds leading to yield loss or yield gain, without re-running the original complex crop models. Our approach is illustrated with three yield datasets simulatedby 19 maize models, 26 wheat models, and 13 rice models. Several statistical models are fitted to thesedatasets, and are then used to analyze the variability of the yield response to [CO2] and temperature.Based on our results, we show that, for wheat, a [CO2] increase is likely to outweigh the negative effectof a temperature increase of +2◦C in the considered sites. Compared to wheat, required levels of [CO2]increase are much higher for maize, and intermediate for rice. For all crops, uncertainties in simulatingclimate change impacts increase more with temperature than with elevated [CO2]. © 2015 Elsevier B.V. All rights reserved. 650 $aARROZ 650 $aCAMBIO CLIMÁTICO 650 $aMAÍZ 650 $aMODELOS ESTADISTICOS 650 $aTRIGO 653 $aClimate change 653 $aCROP MODEL 653 $aEmulator 653 $aMAIZE 653 $aMeta-model 653 $aMODELIZACIÓN DE LOS CULTIVOS 653 $aRICE 653 $aStatistical model 653 $aWHEAT 653 $aYield 700 1 $aASSENG, S. 700 1 $aEWERT, F. 700 1 $aBASSU, S. 700 1 $aDURAND, J.L. 700 1 $aLI, T. 700 1 $aMARTRE, P. 700 1 $aADAM, M. 700 1 $aAGGARWAL, P.K. 700 1 $aANGULO, C. 700 1 $aBARON, C. 700 1 $aBASSO, B. 700 1 $aBERTUZZI, P. 700 1 $aBIERNATH, C. 700 1 $aBOOGAARD, H. 700 1 $aBOOTE, K.J. 700 1 $aBOUMAN, B. 700 1 $aBREGAGLIO, S. 700 1 $aBRISSON, N. 700 1 $aBUIS, S. 700 1 $aCAMMARANO, D. 700 1 $aCHALLINOR, A.J. 700 1 $aCONFALONIERI, R. 700 1 $aCONIJN, J.G. 700 1 $aCORBEELS, M. 700 1 $aDERYNG, D. 700 1 $aDE SANCTIS, G. 700 1 $aDOLTRA, J. 700 1 $aFUMOTO, T. 700 1 $aGAYDON, D. 700 1 $aGAYLER, S. 700 1 $aGOLDBERG, R. 700 1 $aGRANT, R.F. 700 1 $aGRASSINI, P. 700 1 $aHATFIELD, J.L. 700 1 $aHASEGAWA, T. 700 1 $aHENG, L. 700 1 $aHOEK, S. 700 1 $aHOOKER, J. 700 1 $aHUNT, L.A. 700 1 $aINGWERSEN, J. 700 1 $aIZAURRALDE, R.C. 700 1 $aJONGSCHAAP, R.E.E. 700 1 $aJONES, J.W. 700 1 $aKEMANIAN, R.A. 700 1 $aKERSEBAUM, K.C. 700 1 $aKIM, S.-H. 700 1 $aLIZASO, J. 700 1 $aMÜLLER, C. 700 1 $aNAKAGAWA, H. 700 1 $aNARESH KUMAR, S. 700 1 $aNENDEL, C. 700 1 $aO'LEARY, G.J. 700 1 $aOLESEN, J.E. 700 1 $aORIOL, P. 700 1 $aOSBORNE, T.M. 700 1 $aPALOSUO, T. 700 1 $aPRAVIA, V. 700 1 $aPRIESACK, E. 700 1 $aRIPOCHE, D. 700 1 $aROSENZWEIG, C. 700 1 $aRUANE, A.C. 700 1 $aRUGET, F. 700 1 $aSAU, F. 700 1 $aSEMENOV, M.A. 700 1 $aSHCHERBAK, I. 700 1 $aSINGH, B. 700 1 $aSINGH, U. 700 1 $aSOO, H.K. 700 1 $aSTEDUTO, P. 700 1 $aSTÖCKLE, C. 700 1 $aSTRATONOVITCH, P. 700 1 $aSTRECK, T. 700 1 $aSUPIT, I. 700 1 $aTANG, L. 700 1 $aTAO, F. 700 1 $aTEIXEIRA, E.I. 700 1 $aTHORBURN, P. 700 1 $aTIMLIN, D. 700 1 $aTRAVASSO, M. 700 1 $aRÖTTER, R.P. 700 1 $aWAHA, K. 700 1 $aWALLACH, D. 700 1 $aWHITE, J.W. 700 1 $aWILKENS, P. 700 1 $aWILLIAMS, J.R. 700 1 $aWOLF, J. 700 1 $aYIN, X. 700 1 $aYOSHIDA, H. 700 1 $aZHANG, Z. 700 1 $aZHU, Y. 773 $tAgricultural and Forest Meteorology, 2015$gv.214-215, p. 483-493.
Descargar
Esconder MarcPresentar Marc Completo |
Registro original : |
INIA Las Brujas (LB) |
|
Biblioteca
|
Identificación
|
Origen
|
Tipo / Formato
|
Clasificación
|
Cutter
|
Registro
|
Volumen
|
Estado
|
Volver
|
Expresión de búsqueda válido. Check! |
|
|