Optimal Convex Relaxation-based Wavelet Covariance Transform for More Robust AOD-PM Characterization and Tracer Tracking of Biomass Burning Over Land/Sea Boundary Regions

Satellite-Based (SB) Remote Sensing (R/S) column values, such as Aerosol Optical Depth (AOD), and surface values, such as Particulate Matter (PM), can be leveraged for better monitoring and tracking of the pattern of life for peatland fire-induced smoke, smog, and precipitation-generating cloud cover (a Potentially Toxic Triumvirate or PTT), which might initially move out over coastal waters, but subsequently return to coastal lands, peninsulas, as well as nearby islands with ensuing effects. In many cases, as noted by interviews with those involved with firefighting task forces, the tracking of the PTT stops at the border between land/sea. Traditionally, the amalgam of SB R/S and land surface values have been more readily available. However, the availability of Version 3 Maritime Aerosol Network (MAN) data and ship-based measurements of opportunity, such as via the Amazon procurable Microtops II Sunphotometer (MIIS), have made sea surface values more prevalent. Available Moderate Resolution Imaging Spectroradiometer (MODIS) and other SB R/S data can be combined with MIIS data for enhanced contextualization. Conjoined with the fact that aerosols from the peatland fires have been characterized, the Biomass Burning (BB) (e.g., peatland fire) tracer tracking of, among others, Levoglucosan (LG) in PM
2.5
and PM
1
presents an opportunity for both Post-fire Evaluation and Pre-fire Planning. Yet, a core challenge centers upon the issue that while SB R/S estimation of PM
2.5
has become quite pervasive, the estimation of PM
1
requires more care, such as by way of a more robust AOD-PM characterization. Contributory to this goal is AOD normalization, such as by considering Planet Boundary Layer Height (PBLH). Traditionally, this can be derived using a Wavelet Covariance Transform (WCT) (e.g., Haar) for boundary detection in the scattering ratio, and the main contribution of this paper is that of a more robust implementation of the WCT by way of an Optimal Convex Relaxation (OCR)-based approach. The ensuing enhanced AOD-PM characterization, improved AOD normalization, enhanced tracking, and more reliable contextualization might be of value-added proposition for better understanding the impact of peatland fires and improving Postfire Evaluation and Pre-fire Planning.