基于Sentinel-2数据的干旱区典型绿洲植被叶绿素含量估算
|
摘要
以渭干河—库车河绿洲(渭—库绿洲)为研究区,采用在机器学习方面具有明显优势的随机森林回归算法,对绿洲内的4种典型植被(棉花、芦苇、杨树、大枣)叶片的叶绿素相对含量(soil and plant analyzer development,SPAD)进行估算和验证。首先基于"红边"处光谱信息丰富的哨兵2号(Sentinel-2)影像和由其衍生的一阶微分、二阶微分影像各提取23种对叶绿素敏感的宽波段光谱指数,加入3种影响植物生长的土壤参量(土壤含水量,土壤有机质,土壤电导率)作为影响叶片SPAD的特征变量,再根据以上特征变量对每种植被叶片各建立3种方案的SPAD估算模型,从而实现对绿洲内植被叶绿素的监测。结果表明:①影像经一阶微分再提取的植被指数相比原位光谱植被指数,在SPAD估测模型中起到了更重要的作用,在随机森林算法的重要性排序中位居前列;②4种植被叶片的SPAD估测模型都取得了不错的效果,芦苇叶片尤为显著,确定系数(R~2)达到了0. 926;③分析对比3种方案下模型预测能力,方案3(包含土壤参量)的预测能力卓越〔2. 143 <相对百分比偏差(RPD)<2. 692〕,其预测能力排序为:方案3>方案1>方案2,土壤属性和模型预测结果有较强的非线性相关。Sentinel-2数据具有理想的估算绿洲植被叶绿素含量的潜力,提供了一种高效、低成本、潜在高精度的方案来估算叶绿素含量,可为干旱区绿洲农业、生态系统实现更有效的保护和管理提供参考。
The Ogan-Kuqa River Delta Oasis,a typical oasis in the arid zone in China,was taken as the study area. The method of Random Forest with a comparative advantage in machine learning was chosen to model and estimate the relative contents of chlorophyll (SPAD values) of leaves from four kinds of representative vegetation (cotton,reed,poplar and jujube). The 23 broadband spectral indices of vegetation,which are sensitive to chlorophyll content,were obtained based on the reflectance of original Sentinel-2 image with rich spectral information in the"red edge"bands. These vegetation indices were extracted again in the original band order on the firstly-derived Sentinel-2 image and secondly-derived Sentinel-2 image. Three soil parameters (soil moisture content,SMC; soil organic matter,SOM; electrical conductivity,EC) related to vegetation growth were all conducted as the characteristic variables affecting SPAD values. According to the characteristic variables above,three modelling schemes could be developed to monitor the SPAD values of vegetation leaves in oasis. The results showed that: ① Vegetation indices obtained from the firstly-derived image played a more important role than the original vegetation indices in the SPAD estimation model. ② It could be concluded that SPAD-RF regression model,based on the Sentinel-2 satellite image data,could be used to effectively monitor the SPAD values of leaves of the four vegetation types. Especially for the estimation model of SPAD of reed leaves,R~2 reached 0. 926. ③ By analyzing and comparing the model prediction capability under the three schemes,the prediction capability of scheme 3 (including soil parameters) was excellent (2. 143 < relative percentage deviation (RPD) < 2. 692),and the prediction capability was ranked as scheme 3 > scheme 1 > scheme 2. There was a significant nonlinear correlation between the soil properties and the model prediction results. Holistically,Sentinel-2 data has great potential for predicting chlorophyll content of oasis vegetation. This study provided an efficient,low-cost,potentially high-precision solution to estimate SPAD.
The Ogan-Kuqa River Delta Oasis,a typical oasis in the arid zone in China,was taken as the study area. The method of Random Forest with a comparative advantage in machine learning was chosen to model and estimate the relative contents of chlorophyll (SPAD values) of leaves from four kinds of representative vegetation (cotton,reed,poplar and jujube). The 23 broadband spectral indices of vegetation,which are sensitive to chlorophyll content,were obtained based on the reflectance of original Sentinel-2 image with rich spectral information in the"red edge"bands. These vegetation indices were extracted again in the original band order on the firstly-derived Sentinel-2 image and secondly-derived Sentinel-2 image. Three soil parameters (soil moisture content,SMC; soil organic matter,SOM; electrical conductivity,EC) related to vegetation growth were all conducted as the characteristic variables affecting SPAD values. According to the characteristic variables above,three modelling schemes could be developed to monitor the SPAD values of vegetation leaves in oasis. The results showed that: ① Vegetation indices obtained from the firstly-derived image played a more important role than the original vegetation indices in the SPAD estimation model. ② It could be concluded that SPAD-RF regression model,based on the Sentinel-2 satellite image data,could be used to effectively monitor the SPAD values of leaves of the four vegetation types. Especially for the estimation model of SPAD of reed leaves,R~2 reached 0. 926. ③ By analyzing and comparing the model prediction capability under the three schemes,the prediction capability of scheme 3 (including soil parameters) was excellent (2. 143 < relative percentage deviation (RPD) < 2. 692),and the prediction capability was ranked as scheme 3 > scheme 1 > scheme 2. There was a significant nonlinear correlation between the soil properties and the model prediction results. Holistically,Sentinel-2 data has great potential for predicting chlorophyll content of oasis vegetation. This study provided an efficient,low-cost,potentially high-precision solution to estimate SPAD.
引文
[1]Hanelt D.Photosynthesis assessed by chlorophyll fluorescence[J].Bioassays,2018,119:169-198.
[2]Jhanji S,Sekhon N K.Evaluation of potential of portable chlorophyll meter to quantify chlorophyll and nitrogen contents in leaves of wheat under different field conditions[J].Indian Journal of Experimental Biology,2018,56(10):750-758.
[3]Chang S X,Robison D J.Nondestructive and rapid estimation of hardwood foliar nitrogen status using the SPAD-502 chlorophyll meter[J].Forest Ecology and Management,2003,181(3):331-338.
[4]Songsri P,Jogloy S,Holbrook C C,et al.Association of root,specific leaf area and SPAD chlorophyll meter reading to water use efficiency of peanut under different available soil water[J].Agricultural Water Management,2009,96(5):790-798.
[5]Swiader J M,Moore A.SPAD-chlorophyll response to nitrogen fertilization and evaluation of nitrogen status in dryland and irrigated pumpkins[J].Journal of Plant Nutrition,2002,25(5):1 089-1 100.
[6]Yuan S,Goron T L,Huang L,et al.Rice leaf lateral asymmetry in the relationship between SPAD and area-based nitrogen concentration[J].Symmetry,2017,9(6):83.
[7]王娟,韩登武,任岗,等.SPAD值与棉花叶绿素和含氮量关系的研究[J].新疆农业科学,2006,43(3):167-170.[Wang Juan,Han Dengwu,Ren Gang,et al.A study on relation between SPAD value,chlorophyll and nitrogen content in cotton[J].Xinjiang Agricultural Sciences,2006,43(3):167-170.]
[8]邓兴耀,姚俊强,刘志辉.基于GIMMS NDVI的中亚干旱区植被覆盖时空变化[J].干旱区研究,2017,34(1):10-19.[Deng Xingyao,Yao Junqiang,Liu Zhihui.Spatiotemporal dynamic change of vegetation coverage in arid regions in central Asia based on GIMMS NDVI[J].Arid Zone Research,2017,34(1):10-19.]
[9]邹红玉,郑红平.浅述植被“红边”效应及其定量分析方法[J].遥感信息,2010(4):112-116.[Zou Hongyu,Zheng Hongping.The effect and method of quantitative analysis of“Red Edge”of vegetation[J].Remote Sensing Information,2010(4):112-116.]
[10]程乾,黄敬峰,王人潮,等.MODIS植被指数与水稻叶面积指数及叶片叶绿素含量相关性研究[J].应用生态学报,2004,15(8):1 363-1 367.[Chen Qian,Huang Jingfeng,Wang Renchao,et al.Correlation analysis of simulated MODIS vegetation indices and rice leaf area index and leaf chlorophyll content[J].Chinese Journal of Applied Ecology,2004,15(8):1 363-1 367.]
[11]Shou L N,Jia L L,Chen X P,et al.Using high-resolution satellite image to evaluate nitrogen status of winter wheat in the North China plain[J].Journal of Plant Nutrition,2007,30(10):1 669-1 680.
[12]宋晓宇,黄文江,王纪华,等.ASTER卫星遥感影像在冬小麦品质监测方面的初步应用[J].农业工程学报,2006,22(9):148-153.[Song Xiaoyu,Huang Wenjiang,Wang Jihua,et al.Preliminary application of ASTER images in winter wheat quality monitoring[J].Transactions of the CSAE,2006,22(9):148-153.]
[13]谭昌伟,王纪华,赵春江,等.利用Landsat TM遥感数据监测冬小麦开花期主要长势参数[J].农业工程学报,2011,27(5):224-230.[Tan Changwei,Wang Jihua,Zhao Chunjiang,et al.Monitoring wheat main growth parameters at anthesis stage by Landsat TM[J].Transactions of the CSAE,2011,27(5):224-230.]
[14]Delegido J,Verrelst J,Alonso L,et al.Evaluation of sentinel22 rededge bands for empirical estimation of green LAI and chlorophyll content[J].Sensors,2011,11(7):7 063-7 081.
[15]Clevers J G P W,Gitelson A A.Remote estimation of crop and grass chlorophyⅡand nitrogen content using red-edge bands on Sentinel-2 and-3[J].International Journal of Applied Earth Observation and Geoinformation,2013,23:344-351.
[16]Verrelst J,Muoz J,Alonso L,et al.Machine learning regression algorithms for biophysical parameter retrieval:Opportunities for Sentinel-2 and-3[J].Remote Sensing of Environment,2012,118:127-139.
[17]李粉玲,王力,刘京,等.基于高分一号卫星数据的冬小麦叶片SPAD值遥感估算[J].农业机械学报,2015,46(9):273-281.[Li Fenling,Wang Li,Liu Jing,et al.Remote sensing estimation of SPAD value for wheat leaf based on GF-1 data[J].Transactions of the Chinese Society for Agricultural Machinery,2015,46(9):273-281.]
[18]王丹丹,程猛,杨瑞红,等.近20 a渭干河绿洲土壤盐分变化特征[J].干旱区研究,2015,32(6):1 076-1 081.[Wang Dandan,Cheng Meng,Yang Ruihong,et al.Changing characteristics of soil salt in Weigan river oasis for the last 20 years[J].Arid Zone Research,2015,32(6):1 076-1 081.]
[19]康璇,王雪梅,柴仲平.近25 a来渭-库绿洲土地利用/覆被变化及其影响因素[J].水土保持通报,2016,36(5):333-339.[Kang Xuan,Wang Xuemei,Chai Zhongping.Land use/cover changes and Influencing factors in delta oasis of Weigan-Kuqa river during last 25 years[J].Bulletin of Soil and Water Conservation,2016,36(5):333-339.]
[20]丁建丽,瞿娟,孙永猛,等.基于MSAVI-WI特征空间的新疆渭干河-库车河流域绿洲土壤盐渍化研究[J].地理研究,2013,32(2):223-232.[Ding Jianli,Qu Juan,Sun Yongmeng,et al.The retrieval model of soil salinization information in arid region based on MSAVI-WI feature space:A case study of the delta oasis in Weigan-Kuqa watershed[J].Geographical Research,2013,32(2):223-232.]
[21]杨爱霞,丁建丽.新疆艾比湖湿地土壤有机碳含量的光谱测定方法对比[J].农业工程学报,2015,31(18):162-168.[Yang Aixia,Ding Jianli.Comparative assessment of two methods for estimation of soil organic carbon content by Vis-NIR spectra in Xinjiang Ebinur Lake Wetland[J].Transactions of the Chinese Society of Agricultural Engineering,2015,31(18):162-168.]
[22]Green M.A relative error bound for balanced stochastic truncation[J].IEEE Transactions on Automatic Control,1988,33(10):961-965.
[23]Filella I,Penuelas J.The red edge position and shape as indicators of plant chlorophyll content,biomass and hydric status[J].International Journal of Remote Sensing,1994,15(7):1 459-1 470.
[24]Shang J,Liu J,Ma B,et al.Mapping spatial variability of crop growth conditions using RapidEye data in Northern Ontario,Canada[J].Remote Sensing of Environment,2015,168:113-125.
[25]Gitelson A A,Gritz Y,Merzlyak M N.Relationships between leaf chlorophyll content and spectral reflectance and algorithms for nondestructive chlorophyll assessment in higher plant leaves[J].Journal of Plant Physiology,2003,160(3):271-282.
[26]Jiang Z,Huete A R,Didan K,et al.Development of a two-band enhanced vegetation index without a blue band[J].Remote Sensing of Environment,2008,112(10):3 833-3 845.
[27]Haboudane D,Tremblay N,Miller J R,et al.Remote estimation of crop chlorophyll content using spectral indices derived from hyperspectral data[J].IEEE Transactions on Geoscience and Remote Sensing,2008,46(2):423-437.
[28]Vincini M,Frazzi E,D’Alessio P.A broad-band leaf chlorophyll vegetation index at the canopy scale[J].Precision Agriculture,2008,9(5):303-319.
[29]Wu C,Niu Z,Tang Q,et al.Estimating chlorophyll content from hyperspectral vegetation indices:Modeling and validation[J].Agricultural and Forest Meteorology,2008,148(8-9):1 230-1 241.
[30]程志庆,张劲松,孟平,等.杨树叶片叶绿素含量高光谱估算模型研究[J].农业机械学报,2015,46(8):264-271.[Cheng Zhiqing,Zhang Jinsong,Meng Ping,et al.Hyperspectral estimation model of chlorophyll content in poplar leaves[J].Transactions of the Chinese Society for Agricultural Machinery,2015,46(8):264-271.]
[31]葛翔宇,丁建丽,王敬哲,等.基于竞争适应重加权采样算法耦合机器学习的土壤含水量估算[J].光学学报,2018,38(10):1-8.[Ge Xiangyu,Ding Jianli,Wang Jingzhe,et al.Estimation of soil moisture content based on competitive adaptive reweighted sampling algorithm coupled with machine learning[J].Acta Optica Sinica,2018,38(10):1-8.]
[32]Armstrong J S,Collopy F.Error measures for generalizing about forecasting methods:Empirical comparisons[J].International Journal of Forecasting,1992,8(1):69-80.
[33]Main R,Cho M A,Mathieu R,et al.An investigation into robust spectral indices for leaf chlorophyll estimation[J].ISPRS Journal of Photogrammetry and Remote Sensing,2011,66(6):751-761.
[34]Mutanga O,Adam E,Cho M A.High density biomass estimation for wetland vegetation using WorldView-2 imagery and random forest regression algorithm[J].International Journal of Applied Earth Observation and Geoinformation,2012,18:399-406.
[2]Jhanji S,Sekhon N K.Evaluation of potential of portable chlorophyll meter to quantify chlorophyll and nitrogen contents in leaves of wheat under different field conditions[J].Indian Journal of Experimental Biology,2018,56(10):750-758.
[3]Chang S X,Robison D J.Nondestructive and rapid estimation of hardwood foliar nitrogen status using the SPAD-502 chlorophyll meter[J].Forest Ecology and Management,2003,181(3):331-338.
[4]Songsri P,Jogloy S,Holbrook C C,et al.Association of root,specific leaf area and SPAD chlorophyll meter reading to water use efficiency of peanut under different available soil water[J].Agricultural Water Management,2009,96(5):790-798.
[5]Swiader J M,Moore A.SPAD-chlorophyll response to nitrogen fertilization and evaluation of nitrogen status in dryland and irrigated pumpkins[J].Journal of Plant Nutrition,2002,25(5):1 089-1 100.
[6]Yuan S,Goron T L,Huang L,et al.Rice leaf lateral asymmetry in the relationship between SPAD and area-based nitrogen concentration[J].Symmetry,2017,9(6):83.
[7]王娟,韩登武,任岗,等.SPAD值与棉花叶绿素和含氮量关系的研究[J].新疆农业科学,2006,43(3):167-170.[Wang Juan,Han Dengwu,Ren Gang,et al.A study on relation between SPAD value,chlorophyll and nitrogen content in cotton[J].Xinjiang Agricultural Sciences,2006,43(3):167-170.]
[8]邓兴耀,姚俊强,刘志辉.基于GIMMS NDVI的中亚干旱区植被覆盖时空变化[J].干旱区研究,2017,34(1):10-19.[Deng Xingyao,Yao Junqiang,Liu Zhihui.Spatiotemporal dynamic change of vegetation coverage in arid regions in central Asia based on GIMMS NDVI[J].Arid Zone Research,2017,34(1):10-19.]
[9]邹红玉,郑红平.浅述植被“红边”效应及其定量分析方法[J].遥感信息,2010(4):112-116.[Zou Hongyu,Zheng Hongping.The effect and method of quantitative analysis of“Red Edge”of vegetation[J].Remote Sensing Information,2010(4):112-116.]
[10]程乾,黄敬峰,王人潮,等.MODIS植被指数与水稻叶面积指数及叶片叶绿素含量相关性研究[J].应用生态学报,2004,15(8):1 363-1 367.[Chen Qian,Huang Jingfeng,Wang Renchao,et al.Correlation analysis of simulated MODIS vegetation indices and rice leaf area index and leaf chlorophyll content[J].Chinese Journal of Applied Ecology,2004,15(8):1 363-1 367.]
[11]Shou L N,Jia L L,Chen X P,et al.Using high-resolution satellite image to evaluate nitrogen status of winter wheat in the North China plain[J].Journal of Plant Nutrition,2007,30(10):1 669-1 680.
[12]宋晓宇,黄文江,王纪华,等.ASTER卫星遥感影像在冬小麦品质监测方面的初步应用[J].农业工程学报,2006,22(9):148-153.[Song Xiaoyu,Huang Wenjiang,Wang Jihua,et al.Preliminary application of ASTER images in winter wheat quality monitoring[J].Transactions of the CSAE,2006,22(9):148-153.]
[13]谭昌伟,王纪华,赵春江,等.利用Landsat TM遥感数据监测冬小麦开花期主要长势参数[J].农业工程学报,2011,27(5):224-230.[Tan Changwei,Wang Jihua,Zhao Chunjiang,et al.Monitoring wheat main growth parameters at anthesis stage by Landsat TM[J].Transactions of the CSAE,2011,27(5):224-230.]
[14]Delegido J,Verrelst J,Alonso L,et al.Evaluation of sentinel22 rededge bands for empirical estimation of green LAI and chlorophyll content[J].Sensors,2011,11(7):7 063-7 081.
[15]Clevers J G P W,Gitelson A A.Remote estimation of crop and grass chlorophyⅡand nitrogen content using red-edge bands on Sentinel-2 and-3[J].International Journal of Applied Earth Observation and Geoinformation,2013,23:344-351.
[16]Verrelst J,Muoz J,Alonso L,et al.Machine learning regression algorithms for biophysical parameter retrieval:Opportunities for Sentinel-2 and-3[J].Remote Sensing of Environment,2012,118:127-139.
[17]李粉玲,王力,刘京,等.基于高分一号卫星数据的冬小麦叶片SPAD值遥感估算[J].农业机械学报,2015,46(9):273-281.[Li Fenling,Wang Li,Liu Jing,et al.Remote sensing estimation of SPAD value for wheat leaf based on GF-1 data[J].Transactions of the Chinese Society for Agricultural Machinery,2015,46(9):273-281.]
[18]王丹丹,程猛,杨瑞红,等.近20 a渭干河绿洲土壤盐分变化特征[J].干旱区研究,2015,32(6):1 076-1 081.[Wang Dandan,Cheng Meng,Yang Ruihong,et al.Changing characteristics of soil salt in Weigan river oasis for the last 20 years[J].Arid Zone Research,2015,32(6):1 076-1 081.]
[19]康璇,王雪梅,柴仲平.近25 a来渭-库绿洲土地利用/覆被变化及其影响因素[J].水土保持通报,2016,36(5):333-339.[Kang Xuan,Wang Xuemei,Chai Zhongping.Land use/cover changes and Influencing factors in delta oasis of Weigan-Kuqa river during last 25 years[J].Bulletin of Soil and Water Conservation,2016,36(5):333-339.]
[20]丁建丽,瞿娟,孙永猛,等.基于MSAVI-WI特征空间的新疆渭干河-库车河流域绿洲土壤盐渍化研究[J].地理研究,2013,32(2):223-232.[Ding Jianli,Qu Juan,Sun Yongmeng,et al.The retrieval model of soil salinization information in arid region based on MSAVI-WI feature space:A case study of the delta oasis in Weigan-Kuqa watershed[J].Geographical Research,2013,32(2):223-232.]
[21]杨爱霞,丁建丽.新疆艾比湖湿地土壤有机碳含量的光谱测定方法对比[J].农业工程学报,2015,31(18):162-168.[Yang Aixia,Ding Jianli.Comparative assessment of two methods for estimation of soil organic carbon content by Vis-NIR spectra in Xinjiang Ebinur Lake Wetland[J].Transactions of the Chinese Society of Agricultural Engineering,2015,31(18):162-168.]
[22]Green M.A relative error bound for balanced stochastic truncation[J].IEEE Transactions on Automatic Control,1988,33(10):961-965.
[23]Filella I,Penuelas J.The red edge position and shape as indicators of plant chlorophyll content,biomass and hydric status[J].International Journal of Remote Sensing,1994,15(7):1 459-1 470.
[24]Shang J,Liu J,Ma B,et al.Mapping spatial variability of crop growth conditions using RapidEye data in Northern Ontario,Canada[J].Remote Sensing of Environment,2015,168:113-125.
[25]Gitelson A A,Gritz Y,Merzlyak M N.Relationships between leaf chlorophyll content and spectral reflectance and algorithms for nondestructive chlorophyll assessment in higher plant leaves[J].Journal of Plant Physiology,2003,160(3):271-282.
[26]Jiang Z,Huete A R,Didan K,et al.Development of a two-band enhanced vegetation index without a blue band[J].Remote Sensing of Environment,2008,112(10):3 833-3 845.
[27]Haboudane D,Tremblay N,Miller J R,et al.Remote estimation of crop chlorophyll content using spectral indices derived from hyperspectral data[J].IEEE Transactions on Geoscience and Remote Sensing,2008,46(2):423-437.
[28]Vincini M,Frazzi E,D’Alessio P.A broad-band leaf chlorophyll vegetation index at the canopy scale[J].Precision Agriculture,2008,9(5):303-319.
[29]Wu C,Niu Z,Tang Q,et al.Estimating chlorophyll content from hyperspectral vegetation indices:Modeling and validation[J].Agricultural and Forest Meteorology,2008,148(8-9):1 230-1 241.
[30]程志庆,张劲松,孟平,等.杨树叶片叶绿素含量高光谱估算模型研究[J].农业机械学报,2015,46(8):264-271.[Cheng Zhiqing,Zhang Jinsong,Meng Ping,et al.Hyperspectral estimation model of chlorophyll content in poplar leaves[J].Transactions of the Chinese Society for Agricultural Machinery,2015,46(8):264-271.]
[31]葛翔宇,丁建丽,王敬哲,等.基于竞争适应重加权采样算法耦合机器学习的土壤含水量估算[J].光学学报,2018,38(10):1-8.[Ge Xiangyu,Ding Jianli,Wang Jingzhe,et al.Estimation of soil moisture content based on competitive adaptive reweighted sampling algorithm coupled with machine learning[J].Acta Optica Sinica,2018,38(10):1-8.]
[32]Armstrong J S,Collopy F.Error measures for generalizing about forecasting methods:Empirical comparisons[J].International Journal of Forecasting,1992,8(1):69-80.
[33]Main R,Cho M A,Mathieu R,et al.An investigation into robust spectral indices for leaf chlorophyll estimation[J].ISPRS Journal of Photogrammetry and Remote Sensing,2011,66(6):751-761.
[34]Mutanga O,Adam E,Cho M A.High density biomass estimation for wetland vegetation using WorldView-2 imagery and random forest regression algorithm[J].International Journal of Applied Earth Observation and Geoinformation,2012,18:399-406.