耕地土壤有機(jī)質(zhì)與速效氮磷鉀含量高光譜遙感反演研究
[Abstract]:The content of soil organic matter and soil nutrients is of great significance to the growth of supplying crops. The development of hyperspectral remote sensing provides an effective technical approach for the monitoring of regional organic matter and soil nutrients. In order to explore the relationship between soil organic matter and nutrient status and satellite image spectra, Hyperion hyperspectral satellite imagery and farmland fertility survey were used to determine the soil properties (organic matter, alkali-hydrolyzed nitrogen, available potassium). The spectral characteristics of available phosphorus), the spectral inversion model of soil attribute is constructed, and the optimal inversion model of each soil attribute is selected, and the accuracy of the model is evaluated by comparing with the measured value. The main results are as follows: the results show that vegetation in remote sensing images has a large degree of interference with the precision of soil attribute inversion model. The precision of the soil attribute inversion model based on the soil pixel in the image is higher than that of the soil attribute inversion model based on the vegetation pixel. The results of soil pixel spectral sensitivity analysis showed that the content of organic matter had a good response to the Hyperion range of 782.95-813.48 nm. The model with the first derivative of reflectivity has the best fitting accuracy (R2 = 0.777RMSE = 5.31), and the correlation between the model inversion results and the measured values (R2 = 0.809 ~ RMSE = 5.19), which can be used for the rapid determination of the distribution of regional organic matter content. The available phosphorus content has a good response ability to the 1467.33 nm ~ 1 00.29 nm band of Hyperion. The fitting accuracy of the model based on the first derivative of reflectivity is the best (R2 = 0.767RMSE = 19.55), and the correlation between the model inversion results and the measured values (R2 = 0.783rMSE = 9.04), and the fitting accuracy of the model is R2 as follows: The RMSE of 0.314 and 0.405 are 38.06 and 52.47 respectively. It can not be used for the rapid determination of alkali-hydrolyzed nitrogen and available potassium in hyperspectral. For vegetation pixels, only available phosphorus content has a good response to the 1457.23 nm band of Hyperion. The fitting effect of the model established by the ratio index is the best (R2 = 0.304 RMSE is 38.96), and the correlation between the model inversion results and the measured values (R2 = 0.740? RMSE = 16.77), organic matter, The fitting accuracy of the optimal inversion model for alkali-hydrolyzed nitrogen and rapidly available potassium is 0.171 0.196 and 0.163 RMSE respectively, which is 9.1U 39.63 and 79.79, which can not be used for the rapid determination of organic matter, alkali-hydrolyzed nitrogen and available potassium in vegetation covered areas in hyperspectral remote sensing images.
【學(xué)位授予單位】:福建農(nóng)林大學(xué)
【學(xué)位級(jí)別】:碩士
【學(xué)位授予年份】:2017
【分類號(hào)】:S158;S127
【參考文獻(xiàn)】
相關(guān)期刊論文 前10條
1 劉煥軍;張小康;張新樂;武洪峰;金慧凝;于勝男;邱政超;;面向土壤分類的高光譜反射特征參數(shù)模型[J];遙感學(xué)報(bào);2017年01期
2 劉凡;馬玲;楊光;陳建華;馬雪蓮;王海江;;灰漠土土壤全氮含量的高光譜特征分析及估測(cè)[J];新疆農(nóng)業(yè)科學(xué);2017年01期
3 Chang TIAN;Xuan ZHOU;Qiang LIU;Jian-wei PENG;Wen-ming WANG;Zhen-hua ZHANG;Yong YANG;Hai-xing SONG;Chun-yun GUAN;;控釋肥料對(duì)早熟油菜產(chǎn)量、養(yǎng)分吸收和肥料利用率的影響(英文)[J];Journal of Zhejiang University-Science B(Biomedicine & Biotechnology);2016年10期
4 南鋒;朱洪芬;畢如田;;黃土高原煤礦區(qū)復(fù)墾農(nóng)田土壤有機(jī)質(zhì)含量的高光譜預(yù)測(cè)[J];中國農(nóng)業(yè)科學(xué);2016年11期
5 金慧凝;張新樂;劉煥軍;康苒;付強(qiáng);寧東浩;;基于光譜吸收特征的土壤含水量預(yù)測(cè)模型研究[J];土壤學(xué)報(bào);2016年03期
6 王莉雯;衛(wèi)亞星;;濕地土壤全氮和全磷含量高光譜模型研究[J];生態(tài)學(xué)報(bào);2016年16期
7 白由路;;植物營養(yǎng)與肥料研究的回顧與展望[J];中國農(nóng)業(yè)科學(xué);2015年17期
8 許麗麗;李寶林;袁燁城;高錫章;劉海江;董貴華;;2000-2010年中國耕地變化與耕地占補(bǔ)平衡政策效果分析[J];資源科學(xué);2015年08期
9 欒福明;熊黑鋼;王芳;時(shí)卉;王昭國;張芳;王晶晶;;基于小波分析的土壤速效K含量高光譜反演[J];干旱區(qū)地理;2015年02期
10 吳才武;夏建新;段崢嶸;;土壤有機(jī)質(zhì)預(yù)測(cè)性制圖方法研究進(jìn)展[J];土壤通報(bào);2015年01期
相關(guān)博士學(xué)位論文 前5條
1 陳紅艷;土壤主要養(yǎng)分含量的高光譜估測(cè)研究[D];山東農(nóng)業(yè)大學(xué);2012年
2 張婷婷;基于PLS模型的農(nóng)業(yè)土壤成分高光譜遙感反演研究[D];吉林大學(xué);2010年
3 劉偉東;高光譜遙感土壤信息提取與挖掘研究[D];中國科學(xué)院研究生院(遙感應(yīng)用研究所);2002年
4 李裕元;坡地土壤磷素與水分遷移試驗(yàn)研究[D];西北農(nóng)林科技大學(xué);2002年
5 張富倉;土壤-根系統(tǒng)養(yǎng)分遷移機(jī)制及其數(shù)值模擬[D];西北農(nóng)林科技大學(xué);2001年
相關(guān)碩士學(xué)位論文 前5條
1 李曦;基于高光譜遙感的土壤有機(jī)質(zhì)預(yù)測(cè)建模研究[D];浙江大學(xué);2013年
2 嚴(yán)加亮;武夷山不同海拔土壤磷素的空間異質(zhì)性研究[D];福建農(nóng)林大學(xué);2012年
3 劉青;人工神經(jīng)網(wǎng)絡(luò)遙感影像分類系統(tǒng)的設(shè)計(jì)與實(shí)現(xiàn)[D];昆明理工大學(xué);2012年
4 鮑晨光;森林類型遙感分類研究[D];東北林業(yè)大學(xué);2010年
5 李建勇;測(cè)土配方施肥中土壤有效氮鉀測(cè)定方法研究[D];西南大學(xué);2008年
,本文編號(hào):2175392
本文鏈接:http://sikaile.net/kejilunwen/nykj/2175392.html