热红外高光谱测量中非目标因素的影响

Non-target Factor Impacts on Thermal Hyperspectral Measurement

野外热红外高光谱测量对环境背景辐射的扰动极为敏感,目前尚缺少对干扰因素的系统分析研究。为了对各类非目标影响因素进行量化评价,探讨了测量中的误差来源和构成,并采用基于标准光谱库的数值模拟方法开展了定量/半定量分析。结果表明,非目标因素主要有以下5类:发射率/温度分离算法选取、下行辐射波动、肤温变化、仪器热噪声以及平滑度指数类型,其中分离算法对发射率分离精度影响最大,黑体拟合法误差可达0.06,而平滑度指数影响最小,介于0.001~0.003 5之间;各因素影响程度大小依次为:算法选取>下行辐射波动>肤温变化>仪器热噪声>平滑度指数。该文给出了提高测量精度的操作建议,以期为野外热红外高光谱研究提供有益借鉴。

The in-situ measurement of thermal infrared spectrum is sensitive to the fluctuation of environmental radiation, and there is still no systematic analysis for the influences of interference factors. In order to quantitatively evaluate the impacts of non target factor on measurement results, this paper investigated the sources of variable measuring errors, and semi-quantitatively/quantitatively analyzed the influences of those non-target factors by means of numerical simulation, based on the standard spectral library. The results show that, the interfering factors faced with conducting measurement in field include the selection of TES algorithms, fluctuation of down-ward radiation, surface temperature variation, electric noise and the selection of smooth indexes. Of all these factors, the selection of TES algorithms poses the most significant impact on the accuracy of retrieved emissivity, with the black body simulation method reaching an RMSE of 0.06. The selection of smooth indexes is the least significant factor with an RMSE between 0. 001 - 0. 003 5. According to the degree of influence on measuring accuracy, these factors can be listed as follow： selection of TES algorithms〉 fluctuation of down-ward radiation〉 surface temperature variation〉electric noise〉smooth indexes. Meanwhile, this paper also provided some advices beneficial to improving measurement accuracy, and therefore offering salutary reference for in-situ measurement of thermal infrared spectrum.