目的是研究不同钾营养水平春玉米典型生育期叶片的光谱响应,探索叶片内营养成分与叶片光谱反射率的相关性。方法是设置了不同梯度钾处理的盆栽试验,按玉米生育期进行光谱测定和取样分析。结果,通过对不同钾处理间玉米叶片养分含量的差异性分析表明,随着施钾的提高,叶片钾含量差异性达到显著水平。分析不同钾营养水平不同生育时期春玉米叶片光谱反射率与叶片钾含量的相关关系,并建立了喇叭口期利用叶片光谱反射率估测叶片钾含量的数学模型;以及分析了该处理下喇叭口期叶片内水分、叶绿素、氮、磷、钙、镁、锌、锰、铜、铁含量与叶片光谱反射率的相关性。结果表明：不同生育时期叶片钾含量与其光谱反射率的相关关系在光谱维方向存在明显差别,730—930nm和960—1100nm两波段为春玉米喇叭口期评价钾营养状况的敏感波段,光谱变量R767＋R1057,（R767＋R1057） /（logR767＋logR1057）和（R767-R1057） /（logR767-logR1057）均能很好的预测喇叭口期叶片钾含量;该时期叶片内不同成分与光谱反射率相关分析表明：550nm,710nm,950nm三波段处是各个相关曲线的突变点;叶片内各成分间高度相关的,它们的光谱相关曲线趋势也极为一致或对称。

The application of spectralproperties as a diagnostic tool in nutrientdeficiency diagnosis requires furtherun-derstanding of the relationship between spectralproperties and nutrientconcentration in plant tissue. Therefore, to exam-ine the effectof variable potassium fertilization on corn(ZeamaysL. ) leaves and the response on leaf spectral reflec-tance. We designed an experiment to examine the effectvariable amounts of potassium fertilization during four stages of corn growth. Our results showed that the difference in potassium leaf content is significant at the0.05probability level with increasing potassium fertilization. We then examined the correlation between corn leaves spectra reflectance and nu-trient contentas a function ofvariable levels ofpotassium fertilization over the entire corn growing season by potted plant experimen.t Digitalmodelswere constructed to assess potassium content in leaves by utilizing leaves spectral reflectance in thebooting stageOurresults showed thattherewas a correlation between the spectralreflectanceofcorn leaves and cor-responding leaf innermoisture, chlorophylland othernutrientcontent including nitrogen, phosphorous, potassium, calci-um, magnesium, copper, iron, manganese, and zinc. These differences in the spectral reflectance held for the entire growing season.730—930nm and960—1100nm weredefined as sensitivewavebands to assesspotassium nutrition con-dition in booting stage, and spectral variables includingR767+R1057, (R767+R1057)/( logR767+logR1057) and (R767-\nR1057) /(logR767-logR1057) could predictaccurately corn leafpotassium-content in the booting stage. Ouranalysis indica-ted a correlation between various components in the inner leaves and spectralreflectance atthis stage. Inflection pointsoc-curred at410nm,550nm,710nm,950nm in every curve. When correlationsbetween the various componentswerehigh,\ntheir spectra correlation curveswere coincidentor symmetrica.l