摘要
【目的】基于LANDSAT-5像元尺度,分析棉田生长密度监测准确定性的影响因子,并提出改进方法,探索减弱非棉苗背景空间差异的遥感指数,确立棉田生长密度遥感监测的最佳时相,为棉花估产和棉田农作分区管理提供数据支持。【方法】实地调查13块棉田(630hm2),获取了棉田生长密度、经纬度以及播种时间、出苗时间所组成的60个样区数据,每样区3个样点;从播种期到盛花期5个时相的遥感影像提取EVI和DEVI,样本等分为建模数据和模型检验数据;采取分播期和不分播期两种方式分别使用EVI和DEVI建立棉田生长密度估算模型,其决定系数经过显著水平检验后,进行模型估算准确性检验,并将优势模型应用于县域范围的棉田生长密度监测。【结果】分析表明,出苗时间,大、小苗对棉田生长密度的监测影响较大,以中期播种数据为例所建立的分播期估算模型检验结果表明,在6月9日和6月25日每公顷棉田生长密度的估算误差分别为2.05×104株/hm2和2.07×104株/hm2,而采用不分播期估算方式时,两个时相模型的绝对估算误差为2.80×104株/hm2和2.53×104株/hm2。在5月24日,DEVI对土壤等背景的空间差异消除作用得到表现,与使用EVI相比,监测时间从6月9日提前到5月24日;兼顾模型监测的准确性和时相因素,棉花现蕾到开花期是棉田生长密度监测的最佳时段;以新疆建设兵团148团为例,使用优选出的6月9日I式估算方法进行示例监测,区域监测结果可以较好表明不同棉田生长密度的分配比例和空间分布特征。【结论】出苗时间和土壤等背景是影响棉田生长密度监测准确性的主要因素,分播期估算能显著提高监测准确性,DEVI遥感参数可以使监测时间提前,从现蕾期到开花期是棉田生长密度估算的最佳时段,优选模型可以在县级区域应用。
【Objective】 Based on the Landsat-5 cell level, analyzing the factors affecting the estimating veracity, exploring vegetation indexes to clear up the space information difference of the non-cotton background, ascertaining the optimal time to monitoring the existing cotton-seedling density, as a result, the information would be provided for the cotton yield estimation and zones management. 【Method】 Sixty group sample data, consisting of the existing cotton-seedling density, longitude/latitude, sowing time, emergence time, were obtained through investigating the thirteen fields （630 hm^2）, and three sample dot data in every sample area were averaged. EVI and DEVI were picked up from the images of five times from sowing time to full-flowering. And then sixty group sample data were divided into two equal parts to establish and text models. The linear models were established bydata of the middle sowing time and the all three sowing times on the basis of EVI and DEVI, respectively, and the model veracity was tested by RMSE and REPE. At last, the existing cotton-seedling density at the country scale was retrieved by the best model.【Result】 The analysis results showed that the difference of seedling size caused by the different emergence times debased the estimation veracity; and as the sample of the different sowing times, the testing result of the middle sowing time models showed that the absolute error of the existing cotton-seedling number of each hectare on 9 June and 25 June was 2.05×10^4 plants/hm^2 and 2.08×10^4 plants/hm^2, respectively, the absolute error under the three sowing times was 2.80×10^4 plants/hm^2 and 2.53×10^4 plants/hm^2 respectively. DEVI Compared with EVI on 24 May cleared up the effect of the space difference of the non-cotton background to some extent, and then the estimation time advanced from 9 June to 24 May. Giving attention to veracity and time of models, the optimal time monitoring the existing cotton-seedling density was from budding to full-flowering. As an example, I function on 9 June was used to monitor the existing cotton-seedling density in the 148th farm of Xinjiang Construction Crops, the result could exhibit the distributing proportion and space characteristics rightly. 【Conclusion】 The result showed that emergence time and the space background difference are the main factors affecting the estimation veracity of the existing cotton-seedling density, and the models based on different sowing times could improve the estimation level, and DEVI could make the monitoring time in advance, and the optimal time for estimating the existing cotton-seedling density was from budding to full-flowering, and the demonstration indicated that the research result was feasible.
出处
《中国农业科学》
CAS
CSCD
北大核心
2009年第4期1197-1206,共10页
Scientia Agricultura Sinica
基金
国家科技支撑计划项目(2006BAD10A01)
国家“863”项目(2006AA10Z207、2006AA10A302)
关键词
棉花
生长密度
像元
遥感
cotton
existing cotton-seedling density
cell level
monitoring
