Logo image
First multi-sensor characterization of fire-emitted potassium spectral signatures from space in low-light and daytime conditions
Journal article   Peer reviewed

First multi-sensor characterization of fire-emitted potassium spectral signatures from space in low-light and daytime conditions

William Julstrom, Jun Wang and Meng Zhou
Remote sensing of environment, Vol.344, 115542
10/2026
DOI: 10.1016/j.rse.2026.115542

View Online

Abstract

The spectrum of burning biomass contains prominent potassium (K) emission lines at 766.7 nm and 770.1 nm. These emission lines are unique to flaming combustion, potentially allowing their use in development of remote sensing techniques to characterize fire phase, flaming temperature, and allow more accurate estimates of fire emissions. Previous studies have shown modest success in detecting potassium emission lines (K-lines) using airborne instruments or meter-scale spatial resolution satellites. This study, for the first time, shows the detection of K-lines in wildfire spectra under low-light conditions using the TROPOspheric Monitoring Instrument (TROPOMI), and under local noon daylight conditions using the Orbiting Carbon Observatory-2 (OCO-2). Our observations clearly show separation of dual K-line peaks and that the line at 766.7 nm (not 770.1 nm) is attenuated strongly by oxygen absorption, in contrast with previous studies. We also show that K-lines can be isolated from the background signal and atmospheric effects using Principal Component Analysis (PCA) and that the magnitude of the Principal Component (PC) corresponding to the K-lines correlates well with both fire Visible Light Power (VLP) and Fire Radiative Power (FRP) as shown by the VIIRS FIre Light Detection Algorithm (FILDA), with VLP showing better correlation. This suggests quantitatively that K-lines could add critical information to improve existing active fire algorithms for deriving a daytime fire combustion efficiency and other characteristics to better estimate fire emissions. •First detection of potassium lines in wildfire spectra using TROPOMI and OCO-2.•The 770.1 nm emission line is newly found largely free of oxygen A-band absorption.•Principal Component Analysis effectively isolates the potassium emission line signal.•The potassium emission line is stronger in fires containing more flaming combustion.
Biomass burning Hyperspectral imaging MODIS OCO-2 Principal component analysis TROPOMI VIIRS

Details

Metrics

2 Record Views
Logo image