Preflight Spectral Calibration of Airborne Shortwave Infrared Hyperspectral Imager with Water Vapor Absorption Characteristics.

Honglin Liu,Dong Zhang,Yueming Wang

doi:10.3390/s19102259

Abstract

Due to the strong absorption of water vapor at wavelengths of 1350–1420 nm and 1820–1940 nm, under normal atmospheric conditions, the actual digital number (DN) response curve of a hyperspectral imager deviates from the Gaussian shape, which leads to a decrease in the calibration accuracy of an instrument’s spectral response functions (SRF). The higher the calibration uncertainty of SRF, the worse the retrieval accuracy of the spectral characteristics of the targets. In this paper, an improved spectral calibration method based on a monochromator and the spectral absorptive characteristics of water vapor in the laboratory is presented. The water vapor spectral calibration method (WVSCM) uses the difference function to calculate the intrinsic DN response functions of the spectral channels located in the absorptive wavelength range of water vapor and corrects the wavelength offset of the monochromator via the least-square procedure to achieve spectral calibration throughout the full spectral responsive range of the hyper-spectrometer. The absolute spectral calibration uncertainty is ±0.125 nm. We validated the effectiveness of the WVSCM with two tunable semiconductor lasers, and the spectral wavelength positions calibrated by lasers and the WVSCM showed a good degree of consistency.

Highlights

IntroductionAs an image–spectrum merging technology, hyperspectral imaging has been widely used in agriculture, ocean observations, urban planning, disaster monitoring, and many other fields [1,2,3]
As an image–spectrum merging technology, hyperspectral imaging has been widely used in agriculture, ocean observations, urban planning, disaster monitoring, and many other fields [1,2,3].The quantitative retrieval of a target’s surface spectral reflection characteristics is one of the important features of hyperspectral imagers, requiring an accurate spectral position for the instrument
The popular methods for spectral calibration are divided into two main categories: The characteristic spectrum calibration method (CSCM), which relies on sources with unique spectral properties, such as tunable lasers [6], filters containing rare earth oxides [7], atmospheric characteristic absorption lines [8,9], gas molecules absorb cells [10], spectrum lamps [11], etc

Summary

Introduction

As an image–spectrum merging technology, hyperspectral imaging has been widely used in agriculture, ocean observations, urban planning, disaster monitoring, and many other fields [1,2,3]. The monochromatic collimated spectrum calibration method has gradually become the first choice for spectral calibration of spectrometers, but the uncertainty of the monochromator’s stability has always been the bottleneck limiting the accuracy of spectrometers In response to such problems, Zhang et al [15] analyzed the relationship between the mechanical error of the monochromator system and the wavelength of the emitted light and established a mathematical model to calculate the monochromatic light’s wavelength offset. We studied the actual DN response curve in the wavelength range mentioned above and found that each absorption valley along the actual DN response curve of every spectral channel located in the absorptive range corresponds to the spectral absorption characteristics of water vapor This phenomenon helps us to solve the problem of the decrease in the spectral calibration accuracy of the spectral channels, and corrects the wavelength offset of the monochromator simultaneously. The experimental results in this paper were consistent with the theory, which confirmed the feasibility of the water vapor spectrum calibration method

Methods

Section 3.

LSH-T250

Result

Findings

Conclusions