林維潘,李懷民,倪 軍,蔣小平,朱 艷
·農業(yè)信息與電氣技術·
基于便攜式三波段作物生長監(jiān)測儀的水稻長勢監(jiān)測
林維潘,李懷民,倪 軍,蔣小平,朱 艷※
(1. 南京農業(yè)大學農學院,南京 210095;2. 國家信息農業(yè)工程技術中心,南京 210095;3. 教育部智慧農業(yè)工程研究中心,南京 210095;4. 江蘇省物聯(lián)網技術與應用協(xié)同創(chuàng)新中心,南京 210095)
針對現有兩波段光譜儀在實際應用中存在的植被指數單一、生長指標反演精度低等問題,該研究研發(fā)了一款便攜式三波段作物生長監(jiān)測儀CGMD303(Crop-Growth Monitoring and Diagnosis,CGMD)并于2018年7—9月開展了水稻田間試驗研究。結果表明,CGMD303獲取的植被指數與商用儀器ASD FieldSpec HandHeld2呈良好的線性相關關系,同時基于CGMD303構建的水稻生長監(jiān)測模型可以有效預測葉面積指數、葉片干質量、葉片氮質量分數和葉片氮積累量,決定系數分別為0.85、0.72、0.45和0.68,相對均方根誤差分別為0.21、0.32、0.13和0.39。CGMD303可以有效獲取冠層光譜反射率,構建的水稻指標監(jiān)測模型可以精確預測葉面積指數、生物量和氮素指標,可為水稻田間栽培工作提供決策依據。
水稻;監(jiān)測;光譜;模型;生長參數;便攜式監(jiān)測儀
水稻是中國重要的主糧作物,提高水稻生產水平對保障國家糧食安全和農民經濟收入至關重要[1]。隨著農業(yè)栽培技術的發(fā)展和農業(yè)機械化水平的提高,水稻的生產管理也逐漸走向精確化、智能化,而有效獲取水稻的生長信息是精準農業(yè)的重要前提[2]。傳統(tǒng)的實驗室分析方法雖然可以準確地獲取水稻的生長指標及各項農學參數,但需要破壞性取樣,操作繁瑣,耗時較長,污染環(huán)境且在時空尺度上很難滿足實時、快速、無損的要求,無法適用于現代農業(yè)生產。近年來,光譜監(jiān)測技術的發(fā)展使實時、快速、無損地獲取水稻生長信息成為可能[3]。
不同的地物目標對不同波長(或頻率)的電磁波的吸收、反射和透射特性均存在差異,而細胞結構、生物化學成分和形態(tài)學特征共同決定了作物的光譜反射特性,綠色植物在350~2 500 nm范圍內具有典型的反射光譜特征[4]。國內外學者針對作物生長指標與光譜反射率的定量關系展開了一系列研究。李映雪等[5]指出,810 nm附近近紅外波段反射率是作物葉面積指數的最敏感波段。姚霞等[6]研究發(fā)現,小麥的葉片氮質量分數監(jiān)測最佳波段位于紅邊和近紅外波段。馮偉等[7]和Zhang等[8]指出,紅光波段和近紅外波段是生物量指標的重要敏感波段。Zhu等[9]和Chu等[10]研究發(fā)現,近紅外(810 nm)和紅光波段(660 nm)組合的植被指數對水稻和小麥的氮積累量的監(jiān)測效果最好。此外,已有研究表明,三波段植被指數對一些生長指標的監(jiān)測效果要優(yōu)于兩波段植被指數,適當增加光譜儀波段數有利于提高監(jiān)測精度[11-13]。因此,同時配置可見光、紅邊和近紅外波段的多波段作物生長監(jiān)測設備更加符合作物生長監(jiān)測的需要。綜合考慮各生長指標對應的敏感波段,課題組開發(fā)了一款便攜式三波段作物生長監(jiān)測儀CGMD303,該光譜儀可同時獲取660、730和815 nm的波段反射率,針對不同作物的生長指標建模需要,可選擇適宜的植被指數,在實際應用中更具靈活性[14-16]。
本研究通過水稻大田小區(qū)試驗,利用CGMD303作物生長監(jiān)測診斷儀獲取了不同品種、不同生育期水稻冠層反射光譜,并同步采集水稻相應部位葉面積指數,通過分析不同波段下水稻冠層反射光譜的變化特征,探討構建的光譜植被指數與葉面積指數(Leaf Area Index,LAI)、葉片干質量(Leaf Dry Weight,LDW)、葉片氮質量分數(Leaf Nitrogen Content,LNC)和葉片氮積累量(Leaf Nitrogen Accumulation,LNA)的定量關系,建立基于CGMD303的水稻光譜監(jiān)測模型。本研究在此基礎上評價了CGMD303作物生長監(jiān)測診斷儀在水稻田間生產實際中的性能,為水稻生長的實時監(jiān)測提供參考依據。
于2018年7—9月在江蘇省如皋市國家信息農業(yè)工程技術試驗示范基地(32°27′N,120°77′E)開展水稻田間試驗。試驗設置了2個水稻品種,分別為Ⅱ優(yōu) 728和淮稻5號,旱地育秧后移栽;3個氮肥水平,分別為N0(0 kg/hm2)、N1(150 kg/hm2)、N2(360 kg/hm2),按基肥40%、分蘗肥20%、促花肥20%、?;ǚ?0%施入,施基肥時搭配施入P2O5135 kg/hm2和K2O 220 kg/hm2;2個密度處理,分別為D1(行距×株距為 30 cm×15 cm)和D2(行距×株距為50 cm×15 cm)。試驗采用裂區(qū)設計,以品種為主區(qū),氮肥水平和密度處理為副區(qū),12種處理,3次重復,共計36個小區(qū)。小區(qū)的長和寬分別為6 和5 m,小區(qū)面積為30 m2。小區(qū)梗上覆蓋塑料薄膜,每個小區(qū)獨立排灌。其他栽培管理措施同一般高產田。
CGMD303由三波段傳感器、傳感器支架、水平儀等部件組成。其中,多光譜作物生傳感器在結構上分為上行光傳感器和下行光傳感器。上行光傳感器可以接收太陽光在660、730和815 nm波段處的輻射信息;下行光傳感器用于接收作物冠層在對應波段的反射輻射信息,通過對上下行光傳感器的輻射信息處理獲取660、730和815 nm波段的水稻冠層反射率。CGMD303質量較輕,攜帶方便,適合于田間操作,如圖1所示。
1.三波段傳感器 2.傳感器支架 3.處理器 4.屏蔽電纜線 5.水平儀
1.3.1 光譜數據
選擇無風無云的正午進行水稻冠層光譜數據測試,測試的物候期為水稻拔節(jié)期、孕穗期、抽穗期,獲取660、730和815 nm 3個波段處的冠層反射光譜。試驗使用高光譜儀(Analytical Spectral Devices FieldSpec HandHeld 2,ASD)同步獲取作物冠層反射光譜用于對比。每個小區(qū)選擇3個點進行監(jiān)測,最終結果取每個小區(qū)監(jiān)測結果的平均值。
1.3.2 生長指標
在光譜數據測定同期破壞性取樣獲取水稻的生長指標。每個小區(qū)選擇50 cm具有代表性的水稻植株,在實驗室按照植株器官進行分樣。取樣時使用冠層分析儀LAI-2200C同步獲取葉面積指數(Leaf Area Index,LAI),每個小區(qū)選取3個點進行監(jiān)測,取每個小區(qū)監(jiān)測結果的平均值。樣品在105 ℃下殺青30 min,再于80 ℃條件下將植株烘干至恒質量,稱取葉片干質量(Leaf Dry Weight,LDW,g/m2)。樣品粉碎后用凱氏定氮法測定葉片氮質量分數(Leaf Nitrogen Content,LNC,%)。葉片氮積累量(Leaf Nitrogen Accumulation,LNA,g/m2),如式(1)所示
LNA=LNC?LDW (1)
1.4.1 植被指數計算
本研究參考了常見的歸一化植被指數(Normalized Difference Vegetation Index,NDVI)和差值植被指數(Difference Vegetation Index,DVI)的構建形式,在原植被指數的基礎上增加一個波段,使其保留物理特性的同時增加信息量[17]。其中,三波段植被指數-1(Three-band Vegetation Index-1,TVI-1)的構建借鑒了NDVI 的構建原理及形式,以(red+red-edge)代替NDVI中的紅光波段,并將近紅外波段乘以2使其在數值上與(red+red-edge)相當。三波段植被指數-2(Three-band Vegetation Index-2,TVI-2)的構建借鑒了DVI的構建原理及形式,同樣以(red+red-edge)代替DVI中的紅光波段,將近紅外波段乘以2使其在數值上與(red+red-edge)相當。上述植被指數如式(2)和式(3)所示
TVI-1(2?R1-R2-R3)/(2?R1+R2+R3)(2)
TVI-22?R1-R2-R3(3)
式中R1、R2、R3分別為3種波長的冠層反射率。
1.4.2 數據分析
使用EXCEL2016軟件進行數據處理和圖表制作。將獲取的作物冠層光譜信息構建的植被指數與農學參數進行擬合,通過回歸分析構建作物生長監(jiān)測模型。采用決定系數(coefficient of determination,2)和相對均方根誤差(Relative Root Mean Square Error,RRMSE)來綜合評價模型的性能。數據分析如式(4)和式(5)所示
2=SSR/SST (4)
為評價CGMD303獲取水稻植被指數的性能,將CGMD303和ASD獲取的植被指數進行回歸分析。如圖 2所示,CGMD303獲取的TVI-1和TVI-2與商用儀器ASD獲取的對應值的擬合結果均呈線性關系,但不同植被指數的擬合精度存在一定差異。TVI-1的擬合結果優(yōu)于TVI-2,趨勢線接近1∶1線,R達到0.70,RRMSE為0.20,而TVI-2的相關性較低,R為0.25,RRMSE為0.49。該結果反映出CGMD303獲取的660、730和815 nm反射率數值與ASD存在一定差異,因此兩者獲取的TVI-2性質差異也較大,而TVI-1在TVI-2的基礎上除以(2815-730-660),一定程度上抵消了兩者數值上的差異,因此TVI-1的擬合相關性更高。需要注意的是,盡管CGMD303和ASD獲取的TVI-2的性質存在差異,但并不意味著TVI-2的穩(wěn)定性較差,植被指數的性能仍需要與水稻生長指標進行擬合并驗證。
圖2 便攜式三通道作物生長監(jiān)測儀與便攜式地物高光譜儀的植被指數測定值擬合關系
將CGMD303獲取的三波段植被指數TVI-1和TVI-2分別與水稻生長指標LAI、LDW、LNC和LNA進行擬合構建監(jiān)測模型。張洪程等[18]研究表明,水稻秈粳亞種之間在冠層結構、生理生化指標等方面均存在較大差異。因此,本研究將秈稻品種Ⅱ優(yōu)728和粳稻品種淮稻5號獨立構建監(jiān)測模型,結果如圖3所示。
植被指數與水稻LAI的擬合結果較高,TVI-1和TVI-2與各品種呈指數關系,而相同植被指數下Ⅱ優(yōu)728的LAI數值大于淮稻5號。2種植被指數與LAI的擬合結果相近,其中TVI-1與秈、粳亞種LAI擬合的2分別為0.90、0.74,RRMSE分別為0.12、0.21,而TVI-2 的2分別為0.90、0.71,RRMSE分別為0.12、0.21。植被指數與LDW的擬合關系與LAI類似,也表現為Ⅱ優(yōu)728的LAI數值大于淮稻5號。其中,TVI-1與淮稻5號LDW的擬合結果較好,2為0.62,RRMSE為0.21,而TVI-2與Ⅱ優(yōu)728的擬合結果優(yōu)于TVI-1,2為0.74,RRMSE為0.17。不同亞種水稻LNC的性質與LAI和LDW存在較大差異,淮稻5號的數值大于Ⅱ優(yōu)728,TVI-1對兩者的監(jiān)測效果均優(yōu)于TVI-2,2分別為0.70、0.38,RRMSE分別為0.11、0.16。LNA為LDW和LNC相乘獲取的生長指標,同時包含了生物量和氮素的信息,因此秈粳亞種間的差異相比相比其他生長指標較小。其中,TVI-1與兩種品種的擬合相關性均較高,2分別為0.79、0.59,RRMSE分別為0.22、0.31。上述結果表明,秈稻的生物量和LAI指標大于粳稻,而粳稻的氮素指標則大于秈稻,因此獨立建模有利于提高模型的預測精度。
圖3 植被指數與水稻生長指標的擬合關系
對構建的水稻生長指標監(jiān)測模型的預測效果進行驗證,各生長指標的實測值與預測值的擬合關系如圖4所示。TVI-2對LAI、LDW、LNA的預測效果較好,2分別為0.85、0.72、0.68,RRMSE分別為0.21、0.32、0.39,而TVI-1則對LNC的預測效果優(yōu)于TVI-2,2為0.45,RRMSE為0.13。該結果說明,TVI-1更適用于氮素指標的監(jiān)測,而TVI-2則對LAI和生物量等指標更加敏感。
圖4 水稻生長指標監(jiān)測模型驗證結果
國內外研究機構基于作物光譜監(jiān)測技術開發(fā)了一系列作物生長監(jiān)測設備[19-20]。與GreenSeeker等兩波段光譜儀相比,CGMD303豐富的信息量可以選擇對應的最佳植被指數。盡管ASD FieldSpec等高光譜儀可以獲取更多的植被指數,但這些設備獲取的數據冗余過大且操作復雜,因此很難投入實際應用。CGMD303的3個波段經過前人研究篩選,對多種生長指標敏感,操作更加方便,更適合投入到田間實際應用中。
秈稻品種Ⅱ優(yōu)728和粳稻品種淮稻5號在冠層結構和葉片氮質量分數方面均有較大差異。秈稻的植株高大,LAI和生物量指標相對較大,而粳稻葉色更深,LNC等氮素指標更大。Jacquemoud等[21]和Knyazikhin等[22]的研究表明,冠層反射率受葉片生理生化指標和冠層結構2個方面影響。葉片生理生化指標為葉綠素、氮素、鉀素、含水量等指標,而冠層結構指標包括葉面積指數、覆蓋度、葉傾角、葉方位角等指標,這些指標均會對光譜反射率造成不同程度的影響[23-24]。因此,在對水稻LNC進行監(jiān)測的同時難免會受到來自冠層結構的影響,而僅依靠光譜儀無法分辨不同因素對信號的影響,從而形成“同物異譜”和“同譜異物”的現象[25-26]。因此,本研究將水稻品種作為一種先驗信息有效地將目標群體分類從而提升了模型精度,在以后的研究中可以考慮通過先驗信息對不同種植密度、生育期的群體進行進一步分類,降低冠層結構對光譜監(jiān)測的干擾,這要求操作者具備一定農學知識或者在監(jiān)測過程中使用其他傳感器獲取除光譜外的輔助數據[27-28]。
CGMD303對水稻LNC的監(jiān)測能力始終低于其他生長指標。Verstraete等[29]指出,最佳植被指數應當與目標指標敏感性較高,而與其他指標敏感性較低。本研究所構建的植被指數與冠層群體大小指標敏感性較高,在進行LNC監(jiān)測時難免會受到冠層群體大小的影響。薛利紅等[30]指出綠光波段與藍光波段組合構建的比值和歸一化植被指數與水稻葉片氮質量分數呈顯著負相關,預測精度達到80.09%。Tian等[31-32]使用藍光和綠光波段對水稻進行監(jiān)測,發(fā)現綠光新型比值植被指數SR(545,538)估測LNC的2達到0.73,而對LAI的敏感性較低,一定程度上消除了冠層大小對LNC預測的影響;藍光波段構建的三波段植被指數434/(496+401)預測水稻LNC的2達到0.84,普適性也較好。因此,未來多光譜作物生長傳感器的開發(fā)可以考慮選擇綠光、藍光等更多波段組合的傳感器。
1)本研究使用便攜式三波段作物生長監(jiān)測儀CGMD303(Crop-Growth Monitoring and Diagnosis,CGMD)獲取了拔節(jié)期、孕穗期和抽穗期的水稻冠層植被指數并與商用高光譜儀(Analytical Spectral Devices FieldSpec HandHeld 2,ASD)進行了擬合,結果表明,CGMD303與ASD獲取的植被指數呈線性關系,其中TVI-1的擬合結果更好,決定系數(coefficient of determination,2)達到0.70,相對均方根誤差(Relative Root Mean Square Error,RRMSE)為0.20。CGMD303可以有效獲取水稻冠層光譜數據,數據獲取精確、穩(wěn)定。
2)利用CGMD303獲取的三波段植被指數TVI-1和TVI-2可以實現對水稻葉面積指數(Leaf Area Index,LAI)、葉片干質量(Leaf Dry Weight,LDW)、葉片氮質量分數(Leaf Nitrogen Content,LNC)、葉片氮積累量(Leaf Nitrogen Accumulation,LNA)的有效預測,2分別為0.85、0.72、0.45和0.68,RRMSE分別為0.21、0.32、0.13和0.39。CGMD303監(jiān)測精度高、操作簡單、性價比高,可用于水稻田間栽培指導工作。
[1] 朱德峰,程式華,張玉屏,等. 全球水稻生產現狀與制約因素分析[J]. 中國農業(yè)科學,2010,43(3):474-479. Zhu Defeng, Cheng Shihua, Zhang Yuping, et al. Analysis of status and constraints of rice production in the world[J]. 2010, 43(3): 474-479. (in Chinese with English abstract)
[2] 凌啟鴻,張洪程,丁艷鋒,等. 水稻高產技術的新發(fā)展:精確定量栽培[J]. 中國稻米,2005,11(1):3-7. Ling Qihong, Zhang Hongcheng, Ding Yanfeng, et al. Development of rice high-yielding techniques: Precisely fixed quantity planting[J]. 2005, 11(1): 3-7. (in Chinese with English abstract)
[3] 曹衛(wèi)星,朱艷,田永超,等. 作物精確栽培技術的構建與實現[J]. 中國農業(yè)科學,2011,44(19):3955-3969. Cao Weixing, Zhu Yan, Tian Yongchao, et al. Development and implementation of crop precision cultivation technology[J]. Scientia Agricultura Sinica, 2011, 44(19): 3955-3969. (in Chinese with English abstract)
[4] 姚霞. 小麥冠層和單葉氮素營養(yǎng)指標的高光譜監(jiān)測研究[D].南京:南京農業(yè)大學,2009. Yao Xia. Monitoring Nitrogen Status at Canopy and Leaf Scales with Hyperspectral Sensing in Wheat[D]. Nanjing: Nanjing Agricultural University, 2009. (in Chinese with English abstract)
[5] 李映雪,朱艷,戴廷波,等. 小麥葉面積指數與冠層反射光譜的定量關系[J]. 應用生態(tài)學報,2006,17(8):1443-1447. Li Yingxue, Zhu Yan, Dai Tingbo, et al. Quantitative relationships between leaf area index and canopy reflectance spectra of wheat[J]. Chinese Journal of Applied Ecology, 2006, 17(8): 1443-1447. (in Chinese with English abstract)
[6] 姚霞,朱艷,田永超,等. 小麥葉層氮含量估測的最佳高光譜參數研究[J]. 中國農業(yè)科學,2009,42(8):2716-2725. Yao Xia, Zhu Yan, Tian Yongchao, et al. Research of the optimum hyperspectral vegetation indices on monitoring the nitrogen content in wheat leaves[J]. Scientia Agricultura Sinica, 2009, 42(8): 2716-2725. (in Chinese with English abstract)
[7] 馮偉,朱艷,姚霞,等. 基于高光譜遙感的小麥葉干重和葉面積指數監(jiān)測[J]. 植物生態(tài)學報,2009,33(1):34-44. Feng Wei, Zhu Yan, Yao Xia, et al. Monitoring leaf dry weight and leaf area index in wheat with hyperspectral remote sensing[J]. Chinese Journal of Plant Ecology, 2009, 33(1): 34-44. (in Chinese with English abstract)
[8] Zhang Jiayi, Liu Xia, Liang Yan, et al. Using a portable active sensor to monitor growth parameters and predict grain yield of winter wheat[J]. Sensors, 2019, 19(5): 1108-1125.
[9] Zhu Yan, Yao Xia, Tian Yongchao, et al. Analysis of common canopy vegetation indices for indicating leaf nitrogen accumulations in wheat and rice[J]. International Journal of Applied Earth Observation and Geoinformation, 2008, 10(1): 1-10.
[10] Chu Xu, Guo Yongjiu, He Jiaoyang, et al. Comparison of different hyperspectral vegetation indices for estimating canopy leaf nitrogen accumulation in rice[J]. Agronomy Journal, 2014, 106(5): 1911-1920.
[11] Datt B. A new reflectance index for remote sensing of chlorophyll content in higher plants: Tests using eucalyptus leaves[J]. Journal of Plant Physiology, 1999, 154(1): 30-36.
[12] Rodriguez D, Fitzgerald G J, Belford R, et al. Detection of nitrogen deficiency in wheat from spectral reflectance indices and basic crop eco-physiological concepts[J]. Australian Journal of Agricultural Research, 2006, 57(7): 781-789.
[13] Wang Wei, Yao Xia, Yao Xinfeng, et al. Estimating leaf nitrogen concentration with three-band vegetation indices in rice and wheat[J]. Field Crops Research, 2012, 129: 90-98.
[14] 楊建寧,張井超,朱艷,等. 便攜式作物生長監(jiān)測診斷儀性能試驗[J]. 農業(yè)機械學報,2013,44(4):208-212,201. Yang Jianning, Chao Jingchao, Zhu Yan, et al. Experiments on performance of portable plant growth monitoring diagnostic instrument[J]. Transactions of the Chinese Society for Agricultural Machinery, 2013, 44(4): 208-212, 201. (in Chinese with English abstract)
[15] 倪軍,姚霞,田永超,等. 便攜式作物生長監(jiān)測診斷儀的設計與試驗[J]. 農業(yè)工程學報,2013,29(6):150-156. Ni Jun, Yao Xia, Tian Yongchao, et al. Design and experiments of portable apparatus for plant growth monitoring and diagnosis[J]. Transactions of the Chinese Society of Agricultural Engineering (Transactions of the CSAE), 2013, 29(6): 150-156. (in Chinese with English abstract)
[16] Ni Jun, Zhang Jingchao, Wu Rusong, et al. Development of an apparatus for crop-growth monitoring and diagnosis[J]. Sensors, 2018, 18(9): 3129-3149.
[17] 田永超. 基于高光譜遙感的水稻氮素營養(yǎng)參數監(jiān)測研究[D].南京:南京農業(yè)大學,2008. Tian Yongchao. Monitoring Nitrogen Nutrition Parameters with Hyperspectral Remote Sensing in Rice[D]. Nanjing: Nanjing Agricultural University, 2008. (in Chinese with English abstract)
[18] 張洪程,許軻,張軍,等. 雙季晚粳生產力及相關生態(tài)生理特征[J]. 作物學報,2014,40(2):283-300. Zhang Hongcheng, Xu Ke, Zhang Jun, et al. Productivity and eco-physiological characteristics of late Japonica rice in double-cropping system[J]. Acta Agronomica Sinica, 2014, 40(2): 283-300. (in Chinese with English abstract)
[19] Erdle K, Mistele B, Schmidhalter U. Comparison of active and passive spectral sensors in discriminating biomass parameters and nitrogen status in wheat cultivars[J]. Field Crops Research, 2011, 124(1): 74-84.
[20] Hansen P M, Schjoerring J K. Reflectance measurement of canopy biomass and nitrogen status in wheat crops using normalized difference vegetation indices and partial least squares regression[J]. Remote Sensing of Environment, 2003, 86(4): 542-553.
[21] Jacquemoud S, Baret F. PROSPECT: A model of leaf optical properties spectra[J]. Remote Sensing of Environment, 1990, 34(2): 75-91.
[22] Knyazikhin Y, Schull M A, Stenberg P, et al. Hyperspectral remote sensing of foliar nitrogen content[J]. Proceedings of the National Academy of Sciences of the United States of America, 2013, 110(3): 811-812.
[23] Jia Min, Li Wei, Wang Kangkang, et al. A newly developed method to extract the optimal hyperspectral feature for monitoring leaf biomass in wheat[J]. Computers and Electronics in Agriculture, 2019, 165(1): 104942-104948.
[24] Vanderbilt V C, Grant L. Plant canopy specular reflectance model[J]. IEEE Transactions on Geoscience and Remote Sensing, 1985, 23(5): 722-730.
[25] Ustin S L. Remote sensing of canopy chemistry[J]. Proceedings of the National Academy of Sciences, 2013, 110(3): 804-805.
[26] Townsend P A, Serbin S P, Kruger E L, et al. Disentangling the contribution of biological and physical properties of leaves and canopies in imaging spectroscopy data[J/OL]. Proceedings of the National Academy of Sciences, 2013, 110(12), [2013-03-19], https: //doi. org/10. 1073/pnas. 1300952110.
[27] Darvishzadeh R, Skidmore A, Atzberger C, et al. Estimation of vegetation LAI from hyperspectral reflectance data: Effects of soil type and plant architecture[J]. International Journal of Applied Earth Observation and Geoinformation, 2008, 10(3): 358-373.
[28] Peng Yi, Nguy-Robertson A, Arkebauer T, et al. Assessment of canopy chlorophyll content retrieval in maize and soybean: Implications of hysteresis on the development of generic algorithms[J]. Remote sensing, 2017, 9(3): 226-241.
[29] Verstraete M M, Pinty B. Designing optimal spectral indexes for remote sensing applications[J]. IEEE Transactions on Geoscience and Remote Sensing, 1996, 34(5): 1254-1265.
[30] 薛利紅,楊林章,范小暉. 基于碳氮代謝的水稻氮含量及碳氮比光譜估測[J]. 作物學報,2006,32(3):430-435. Xue Lihong, Yang Linzhang, Fan Xiaohui. Estimation of nitrogen content and C/N in rice leaves and plant with canopy reflectance spectra[J]. Acta Agronomica Sinica, 2006, 32(3): 430-435. (in Chinese with English abstract)
[31] Tian Yongchao, Yao Xia, Yang Jie, et al. Assessing newly developed and published vegetation indices for estimating rice leaf nitrogen concentration with ground- and space-based hyperspectral reflectance[J]. Field Crops Research, 2011, 120(2): 299-310.
[32] Tian Yongchao, Gu Kaijian, Chu Xu, et al. Comparison of different hyperspectral vegetation indices for canopy leaf nitrogen concentration estimation in rice[J]. Plant and Soil, 2014, 376(1/2): 193-209.
Monitoring rice growth based on a portable three-band instrument for crop growth monitoring and diagnosis
Lin Weipan, Li Huaimin, Ni Jun, Jiang Xiaoping, Zhu Yan※
(1.,,210095,; 2.,210095,; 3.,,210095,; 4.,210095,)
The development of instruments for monitoring and diagnosing crop growth quickly and non-destructively obtain crop growth information, which is very helpful for the production and management of crop fields.Aimed at the problems with the existing two-band instrument used for crop growth monitoring and diagnosis, such as relying on a single vegetation index and low accuracy of growth index retrieval, this study developed a portable three-band instrument for crop-growth monitoring and diagnosis CGMD303 (Crop-Growth Monitoring and Diagnosis, CGMD).The CGMD303 instrument consisted of a multi-spectral crop growth sensor, processor system, sensor holder, level, shielded cable, and other components. Multi-spectral crop sensors were divided into upward light sensors and downward light sensors in structure. The upward light sensor could receive solar radiation information in the 660, 730, and 815 nm bands; the downward light sensor consisted of three detector lenses, which were used to detect the characteristic wavelengths of 660, 730, and 815 nm, respectively. The radiation information would be processed after being converted to electrical signals through the photoelectric detector. To test the monitoring performance of CGMD303 on rice growth, rice field experiments were conducted from July 2018 to September 2018 at the demonstration base of the National Engineering and Technology Center for Information Agriculture in Rugao City, Jiangsu Province, China (32°27′N, 120°77′E), and 2 varieties (Liangyou 728 and Huaidao NO.5), 3 nitrogen levels (0, 150 and 360 kg/hm2) and 2 planting density levels (30 cm×15 cm and 50 cm×15 cm) were set in the rice experiments. The canopy spectral reflectance of 660, 730, and 815 nm was obtained at the jointing stage, booting stage, and heading stage of rice and 2 new three-band vegetation indices were constructed. Fitting results of the vegetation indices obtained by CGMD303 and the commercial instrument ASD FieldSpec HandHeld2 showed a good linear correlation, indicating that CGMD303 effectively obtained rice canopy reflectance. Two three-band vegetation indices obtained by CGMD303 and rice growth parameters were fitted to construct rice growth monitoring models. The highest coefficient of determination values of the three-band vegetation indices and leaf area index, leaf dry weight, leaf nitrogen content, and leaf nitrogen accumulation of indica rice were 0.90, 0.74, 0.70, 0.79, respectively, and the relative root mean square error was 0.12, 0.17, 0.11, 0.22, respectively;the highest coefficient of determination values of the three-band vegetation indices and corresponding growth indices of Japonica rice were 0.74, 0.62, 0.38, 0.59, respectively, and the relative root mean square error was 0.21, 0.21, 0.16, 0.31, respectively. The prediction accuracy of the rice growth monitoring models based on CGMD303 for each growth parameters was tested and the coefficient of determination of leaf area index, leaf dry weight, leaf nitrogen content, and leaf nitrogen accumulation of rice were 0.85, 0.72, 0.45, 0.68, respectively, and the relative root mean square error were 0.21, 0.32, 0.13, 0.39, respectively.Verification results showed that CGMD303 could accurately predict leaf area index, biomass, and nitrogen indices of rice.CGMD303 had the advantages of accurate and stable data acquisition, simple operation, high-cost performance, etc. It was suitable for field operations and had high application potential.
rice; monitoring; spectra; models; growth parameters; portable monitor
林維潘,李懷民,倪軍,等. 基于便攜式三波段作物生長監(jiān)測儀的水稻長勢監(jiān)測[J]. 農業(yè)工程學報,2020,36(20):203-208.doi:10.11975/j.issn.1002-6819.2020.20.024 http://www.tcsae.org
Lin Weipan, Li Huaimin, Ni Jun, et al. Monitoring rice growth based on a portable three-band instrument for crop growth monitoring and diagnosis[J]. Transactions of the Chinese Society of Agricultural Engineering (Transactions of the CSAE), 2020, 36(20): 203-208. (in Chinese with English abstract) doi:10.11975/j.issn.1002-6819.2020.20.024 http://www.tcsae.org
2020-07-01
2020-09-06
國家重點研發(fā)計劃(2017YFD0201501);江蘇省六大人才高峰項目(XYDXX-049);江蘇省重點研發(fā)計劃(BE2018399)
林維潘,主要從事農學信息工程領域的研究。Email:2018814037@njau.edu.cn
朱艷,教授,博士生導師,主要從事農業(yè)信息技術領域的研究。Email:yanzhu@njau.edu.cn
10.11975/j.issn.1002-6819.2020.20.024
S237,TP73
A
1002-6819(2020)-20-0203-06