农业工程学报
農業工程學報
농업공정학보
2014年
24期
141-150
,共10页
农作物%遥感%监测%冬小麦%叶面积指数%高光谱遥感%监测模型
農作物%遙感%鑑測%鼕小麥%葉麵積指數%高光譜遙感%鑑測模型
농작물%요감%감측%동소맥%협면적지수%고광보요감%감측모형
crops%remote sensing%monitoring%winter wheat%leaf area index%hyperspectral remote sensing%monitoring model
高光谱遥感能快速无损获取植被冠层信息，是实现作物长势实时监测的重要技术。为研究不同氮磷水平下冬小麦不同生育时期叶面积指数高光谱遥感监测模型，提高叶面积指数高光谱监测精度，该研究连续5a定位测定黄土高原旱地不同氮磷水平和不同冬小麦品种各生育时期冠层光谱反射率与叶面积指数，通过相关分析、回归分析等统计方法，构建不同生育时期冬小麦叶面积指数监测模型。结果表明：不同氮磷水平下，冬小麦叶面积指数随施肥量增加呈递增趋势，随生育时期改变呈抛物线趋势变化；随着氮磷供应量的增加，冠层光谱反射率在可见光波段显著降低2%～5%（P<0.05），在近红外波段显著增加4%～10%（P<0.05）；不同生育时期叶面积指数与优化土壤调整植被指数、增强型植被指数Ⅱ、新型植被指数、修正归一化差异植被指数、修正简单比值植被指数均达极显著相关（P<0.01）；拔节期、孕穗期、抽穗期、灌浆期和成熟期叶面积指数分别与优化土壤调整植被指数、增强型植被指数Ⅱ、增强型植被指数Ⅱ、修正归一化差异植被指数和修正简单比值植被指数拟合效果较好，决定系数分别为0.952、0.979、0.989、0.960和0.993；以不同年份独立数据验证模型表明，所建预测模型均有较好的验证结果，相对误差分别为13.0%、13.5%、12.8%、12.6%和14.0%，均方根误差分别为：0.313、0.336、0.316、0.316、0.324。因此，优化土壤调整植被指数、增强型植被指数Ⅱ、增强型植被指数Ⅱ、修正归一化差异植被指数和修正简单比值植被指数能有效评价拔节期、孕穗期、抽穗期、灌浆期和成熟期冬小麦叶面积指数。同时，叶面积指数分段监测模型较统一监测模型精度有所改善。该结果为实现不同肥力水平下冬小麦不同生育时期长势精确监测提供理论依据和技术支撑。
高光譜遙感能快速無損穫取植被冠層信息，是實現作物長勢實時鑑測的重要技術。為研究不同氮燐水平下鼕小麥不同生育時期葉麵積指數高光譜遙感鑑測模型，提高葉麵積指數高光譜鑑測精度，該研究連續5a定位測定黃土高原旱地不同氮燐水平和不同鼕小麥品種各生育時期冠層光譜反射率與葉麵積指數，通過相關分析、迴歸分析等統計方法，構建不同生育時期鼕小麥葉麵積指數鑑測模型。結果錶明：不同氮燐水平下，鼕小麥葉麵積指數隨施肥量增加呈遞增趨勢，隨生育時期改變呈拋物線趨勢變化；隨著氮燐供應量的增加，冠層光譜反射率在可見光波段顯著降低2%～5%（P<0.05），在近紅外波段顯著增加4%～10%（P<0.05）；不同生育時期葉麵積指數與優化土壤調整植被指數、增彊型植被指數Ⅱ、新型植被指數、脩正歸一化差異植被指數、脩正簡單比值植被指數均達極顯著相關（P<0.01）；拔節期、孕穗期、抽穗期、灌漿期和成熟期葉麵積指數分彆與優化土壤調整植被指數、增彊型植被指數Ⅱ、增彊型植被指數Ⅱ、脩正歸一化差異植被指數和脩正簡單比值植被指數擬閤效果較好，決定繫數分彆為0.952、0.979、0.989、0.960和0.993；以不同年份獨立數據驗證模型錶明，所建預測模型均有較好的驗證結果，相對誤差分彆為13.0%、13.5%、12.8%、12.6%和14.0%，均方根誤差分彆為：0.313、0.336、0.316、0.316、0.324。因此，優化土壤調整植被指數、增彊型植被指數Ⅱ、增彊型植被指數Ⅱ、脩正歸一化差異植被指數和脩正簡單比值植被指數能有效評價拔節期、孕穗期、抽穗期、灌漿期和成熟期鼕小麥葉麵積指數。同時，葉麵積指數分段鑑測模型較統一鑑測模型精度有所改善。該結果為實現不同肥力水平下鼕小麥不同生育時期長勢精確鑑測提供理論依據和技術支撐。
고광보요감능쾌속무손획취식피관층신식，시실현작물장세실시감측적중요기술。위연구불동담린수평하동소맥불동생육시기협면적지수고광보요감감측모형，제고협면적지수고광보감측정도，해연구련속5a정위측정황토고원한지불동담린수평화불동동소맥품충각생육시기관층광보반사솔여협면적지수，통과상관분석、회귀분석등통계방법，구건불동생육시기동소맥협면적지수감측모형。결과표명：불동담린수평하，동소맥협면적지수수시비량증가정체증추세，수생육시기개변정포물선추세변화；수착담린공응량적증가，관층광보반사솔재가견광파단현저강저2%～5%（P<0.05），재근홍외파단현저증가4%～10%（P<0.05）；불동생육시기협면적지수여우화토양조정식피지수、증강형식피지수Ⅱ、신형식피지수、수정귀일화차이식피지수、수정간단비치식피지수균체겁현저상관（P<0.01）；발절기、잉수기、추수기、관장기화성숙기협면적지수분별여우화토양조정식피지수、증강형식피지수Ⅱ、증강형식피지수Ⅱ、수정귀일화차이식피지수화수정간단비치식피지수의합효과교호，결정계수분별위0.952、0.979、0.989、0.960화0.993；이불동년빈독립수거험증모형표명，소건예측모형균유교호적험증결과，상대오차분별위13.0%、13.5%、12.8%、12.6%화14.0%，균방근오차분별위：0.313、0.336、0.316、0.316、0.324。인차，우화토양조정식피지수、증강형식피지수Ⅱ、증강형식피지수Ⅱ、수정귀일화차이식피지수화수정간단비치식피지수능유효평개발절기、잉수기、추수기、관장기화성숙기동소맥협면적지수。동시，협면적지수분단감측모형교통일감측모형정도유소개선。해결과위실현불동비력수평하동소맥불동생육시기장세정학감측제공이론의거화기술지탱。
Hyperspectral remote sensing can rapidly and nondestructively acquire vegetation canopy information. It is an important real time technology to monitor and manage crop growth. Leaf area index (LAI) is a key parameter for crop growth evaluation and yield prediction. The objectives of this study were to establish wheat LAI estimation model based on winter wheat (Triticum aestivum L.) canopy hyperspectral reflectance with different rates of nitrogen or phosphorus application, and to improve the forecast precision of the LAI estimation model at different growth stages of winter wheat on the Loess Plateau of China. The experiments were carried out during 2009-2014 at Northwest A&F University, Yangling, China. The treatments included different winter wheat varieties with various drought resistance grown under five nitrogen fertilizer application rates (0, 75, 150, 225 and 300 kg N/hm2) and four phosphorus application rates (0, 60, 120 and 180 kg P2O5/hm2) LAI and canopy hyperstpectral reflectance of different varieties under fertilizer treatments were measured at jointing, booting, heading, grain filling and maturity stage, respectively. Then LAI monitoring models at different growth stages of winter wheat were constructed by using correlation analysis, regression analysis. The results showed that LAI of wheat was increased with increase in nitrogen and phosphorus application rate at different growth stages, and LAI from jointing to maturity showed a parabolic curve, and the maximum LAI of wheat was at heading stage. When nitrogen or phosphorus application was sufficient, the canopy hyperspectral reflectance of wheat was reduced 3%-5% in the visible wavelength (P<0.05), and increased 4%-10% in the near infrared wavelength (P<0.05). There were significant (P<0.01) correlation between Optimized Soil Adjusted Vegetation Index (OSAVI), Enhanced Vegetation Index Ⅱ (EVI2), New Vegetation Index (NVI), Modified Normalized Difference Vegetation Index (mNDVI), and Modified Simple Ratio Index (mSRI) to LAI, the range of correlation coefficient were from 0.852 to 0.987 at different growth stages. Monitoring models based on OSAVI, EVI2, EVI2, mNDVI and mSRI produced better estimation for LAI at jointing, booting, heading, grain filling and maturity respectively, and R2 were respectively 0.952, 0.979, 0.989, 0.960 and 0.993 (P<0.01). Meanwhile, compared the predicted value and measured value to verify reliability and applicability of the model, relative error of the measured value and predicted value were 13.0%, 13.5%, 12.8%, 12.6% and 0.14.0%, and root mean square error were 0.313, 0.336, 0.316, 0.316 and 0.324, at jointing, booting, heading, grain filling and maturity stage, respectively. Therefore, vegetation indices of OSAVI, EVI2, EVI2, mNDVI and mSRI was the most suitable indeces for monitoring winter wheat LAI at jointing, booting, heading, grain filling and maturity, respectively. There was high prediction precision with different vegetation indices monitoring LAI of winter wheat at different growth stages. These conclusions has important implications for monitoring LAI of winter wheat in large area on the Loess Plateau. Meanwhile, there is a high prediction accuracy of monitoring model based on the different vegetation indices at different growth stages of winter wheat.This result provides technical support for growth monitoring of winter wheat at different fertility and different growth stages for farmers..
