A Novel Detection Algorithm for Impulse Spike Interferences in Cross-Track Infrared Sounder Short-Wave Spectral Radiance

Xin Jin,Yong Chen,Flavio Iturbide-Sanchez,Banghua Yan,Ninghai Sun

doi:10.1109/jstars.2021.3110257

Abstract

Due to deficiencies in the operational spike detection, many moderate and strong impulse spike interferences constantly remain nondetected in the cross-track infrared sounder (CrIS) sensor data record (SDR) data at the short-wave infrared (SWIR) band over the South Atlantic Anomaly (SAA) region. This study characterizes imaginary radiance spectrum features in the presence of four intensities of spikes at three different hitting locations through radiative transfer model simulations. It is found that periodic ringing patterns appear in the imaginary spectral radiance. Following it, we establish a novel spike detection algorithm based on the spiky feature of power spectral density (PSD) in detrended and normalized imaginary SWIR CrIS spectral radiances. The performance of the algorithm is assessed by applying it to the multiple years of the SNPP and NOAA-20 CrIS full spectral resolution (FSR) radiance data. Approximate 300 and 500 spike events per day are detected for SNPP and NOAA-20 CrIS individually. The spikes can obviously increase the uncertainty in brightness temperature measurements at SWIR window channels that is determined by hitting location of the interferogram arms and the impulse intensities. Strong impulse spikes can trigger spectral radiance uncertainty by several times in SWIR window channels. Even the weakest impulse event can increase spectral radiance uncertainty by 4% in SWIR window channels. In addition, our simulations and one case study of CrIS data show that the spikes can degrade the quality of the CrIS data at the SWIR band with the largest impact at absorbing channels and the smallest at the window channels. The resultant brightness temperature change are 3 K for absorbing channels and up to 1 K for window channels in the SWIR band, demonstrating the significance flagging the CrIS data that are contaminated by the moderate and strong spikes that are missed in the operational data processing.

Highlights

THE Cross-track Infrared Sounder (CrIS) is one of instruments onboard the Suomi National Polar-orbiting Partnership (SNPP) and National Oceanic and Atmospheric Administration (NOAA)-20 satellites
We present a method for detecting spike-contaminated data using Short-Wave infrared (SWIR) CrIS imaginary spectral radiance because the spike-triggered current is of the similar magnitude of SWIR detector current [10]
A simulation is firstly conducted by using a line-by-line radiative transfer model to understand characteristics of imaginary spectral radiance on the SWIR spectrum in the presence of various spikes with different intensities and hitting locations

Summary

INTRODUCTION

THE Cross-track Infrared Sounder (CrIS) is one of instruments onboard the Suomi National Polar-orbiting Partnership (SNPP) and National Oceanic and Atmospheric Administration (NOAA)-20 satellites It provides hyper-spectral radiances of either 1305, i.e., normal spectral resolution (NSR), or 2211, i.e., full spectral resolution (FSR) frequencies emitted from the earth surface and its atmosphere, covering three infrared bands: long-wave (LWIR), mid-wave (MWIR) and short-wave (SWIR). Can be detected using the bit-trim mask table in the onboard data processing, but such strong spike events are rare [9] Another impulse spike detection method is developed in the ground data calibration system by comparing two sides of the interferogram around the zero path difference (ZPD) since the spike destroys its symmetry feature [10].

CHARACTERISTICS OF SPIKE-CONTAMINATED EARTH SCENE IMAGINARY RADIANCE

APPLICATION OF THE ALGORITHM TO THE CRIS DATA

Detection of Spike-Contaminated CrIS Data

Impacts on Quality of CrIS Data at the SWIR Band

Findings

SUMMARY AND CONCLUSION