模拟太阳辐射减弱对不同播期冬小麦田N_(2)O排放影响及光谱估算

Effects of Sowing Dates on N_(2)O Emissions from Winter Wheat Fields under Different Shading Conditions

太阳辐射减弱是全球气候变化的主要特征之一。太阳辐射减弱对农田氧化亚氮(N_(2)O)排放有何影响以及能否基于高光谱遥感进行估算,尚未见报道。通过田间模拟试验,探讨了不同遮阴强度下不同播期对冬小麦田N_(2)O排放的影响及基于冠层高光谱变量估算N_(2)O排放通量。采用普通黑色遮阳网覆盖冬小麦植株冠层,以模拟太阳辐射减弱。遮阴强度设3个水平,即对照(CK,不遮阴,遮阴率为0)、中度遮阴(S1,遮阴率为61.85%)和重度遮阴(S2,遮阴率为83.91%);播期设2个水平,即常规播期(T)和晚播(L)。结果表明:①遮阴显著提高了冬小麦农田N_(2)O排放量,即S2>S1>CK;晚播明显降低麦田N_(2)O排放量。②冬小麦冠层光谱反射率与N_(2)O排放通量的敏感波段,主要在红光波段和近红外波段,拔节期呈显著正相关,最大相关系数为0.92,孕穗—抽穗期呈显著负相关,最大负相关为-0.75。6种植被指数与N_(2)O排放通量呈显著相关性,其中NLI(非线性植被指数)与N_(2)O排放通量的相关系数最佳,为0.93(拔节期)、-0.77(孕穗—抽穗期)。③于拔节期建立了以NLI和629、630 nm原始反射率(ρ629、ρ630)为参数估算N_(2)O排放通量的逐步回归模型,建模集和验证集决定系数R^(2)分别为0.89和0.83;于孕穗—抽穗期建立了以ρ1072、ρ760为参数估算N_(2)O排放通量的双变量模型,R^(2)分别为0.70和0.75。研究认为,通过构建高光谱遥感参数模型估算农田N_(2)O排放量是可行的,这为进一步开展区域农田N_(2)O排放量遥感监测提供了试验参考依据。

Solar radiation reduction is one of the main features of global climate change.Few researches have been available regarding the impact of reduced solar radiation on Nitrous Oxide(N_(2)O)emissions and its estima⁃tion based on hyperspectral remote sensing under field conditions.Field simulation experiment was conducted to investigate the effects of sowing date on N_(2)O emission in winter wheat field under different shading intensities and N_(2)O emission flux estimation based on canopy hyperspectral variables.Ordinary black sunshade net was used to cover the canopy of winter wheat plants to simulate the decrease of solar radiation.The shading intensity was set at three levels,namely,control(CK,no shading,0%shading rate),moderate shading(S1,61.85%shading rate)and heavy shading(S2,83.91%shading rate).The sowing date was set at two levels,namely,traditional sowing date(T)and late sowing date(L).The results showed that:(1)shading significantly in⁃creased N_(2)O emission,i.e.S2>S1>CK;but late sowing significantly reduced N_(2)O emission in winter wheat field.(2)The sensitive bands between spectral reflectance on winter wheat canopy and N_(2)O flux are mainly in the red band and near-infrared band.A significant positive correlation existed at the elongation stage,with the maximum correlation coefficient of 0.92.A significant negative correlation occurred at booting-heading stage,with the maximum negative correlation of-0.75.The significant correlations were observed between six plant⁃ing cover indexes and N_(2)O fluxes,and the highest correlation coefficient was found between NLI(nonlinear vegetation index)and N_(2)O flu,with being 0.93(jointing stage)and-0.77(booting to heading stage).(3)At the elongation stage,a stepwise regression model was established to estimate N_(2)O flux using NLI and original reflectance(ρ629、ρ630)at 629 nm and 630 nm.The coefficient of determination(R^(2))of modeling set and verifica⁃tion set was 0.89 and 0.83,respectively.At booting and heading stage,a bivariate model was established to esti⁃mate N_(2)O flux using the parameters(ρ1072、ρ760),with R^(2)being 0.70 and 0.75,respectively.This study suggests that it is feasible to estimate N_(2)O emission in winter wheat field by hyperspectral remote sensing parameter mod⁃el,and to provide experimental reference for further monitoring regional N_(2)O emissions with remote sensing.