Using point source imaging satellite observations to guide landfill methane model improvements at the national and sub-national scale

As national and sub-national governments, companies, and communities plan methane mitigation action, there is a need for robust emissions tracking systems, especially for major sectors like waste where countries have made commitments to reduce emissions. Landfills are a major source of methane emissions in many jurisdictions spread across the world, so there is a need in the waste sector for monitoring frameworks that are applicable at scale but also provide facility-level insights to guide decision making. 

 

Given the complexity of landfill emissions both in terms of variability and underlying causes, models are a common tool used for planning and tracking landfill methane mitigation, but past studies show potential biases in models and inventories compared to observations. In this work, we bring together both process-level insights as provided in bottom-up models and our top-down observations from the Tanager-1 satellite by (1) improving the accuracy and consistency of satellite-derived annual average emission rates and (2) developing methodologies for reconciling the two unique datasets. The goal of this work is to use satellite methane observations to identify improved bottom-up model parameters, focusing on the modeling frameworks used by national and sub-national jurisdictions.

 

As a point source imaging satellite, Tanager-1 is well suited for tracking emissions at landfills as it provides facility-scale methane emissions data, but existing algorithms and workflows for creating the emissions data have been primarily validated based on controlled release experiments which mimic environments more similar to the oil and gas sector than landfills. We identify methods that are robust and best suited to landfills by performing sensitivity tests for our quantification methods, testing algorithms and parameters, and identifying causes of bias unique to landfill environments (e.g., albedo, topography). The next step is translating our Tanager-1 observations to annual averages. We present a new methodology for temporally averaging satellite observations that accounts for null detects through scene-specific probability of detection limits. Finally, we compare our annual average satellite-based emission estimates to bottom-up models typically used by jurisdictions for official reporting (e.g., IPCC, LandGEM, US GHGRP), focusing on select countries where there is sufficient spatiotemporal coverage with Tanager-1. We use statistical methods to adjust parameters in the bottom-up models to reconcile the model estimates with observed emissions, allowing region-specific model parameter adjustments to account for potential climatic and meteorological factors. Finally, we discuss the implications of our initial results in terms of improvements to official national reporting and compare to inverse modeling results.