Description

This training covers the theoretical approaches and key datasets for estimating the near-surface concentration of particulate matter smaller than 2.5 micrometers in diameter (PM_2.5) by using satellite aerosol optical depth (AOD) and other data sources (including ground-based monitors and model estimates). 

Participants will acquire knowledge about different methodologies, and their strengths and limitations, for inferring surface PM_2.5 from satellite AOD. They will be able to identify access methods for using existing datasets of surface PM_2.5 estimated from satellite sensors and other data sources. They will also learn how to compare satellite-based estimates of PM_2.5 to in situ measurement data through hands-on coding exercises using a generic code template and suitable co-location methodology. This would be the first step in evaluating the suitability of these datasets for the users’ applications and regions.

Prerequisites

Part 3 of this training will require you to access Google Colab; if you are unable to access using the link here, you may need to sign up for a free Google account. Furthermore, you will need a free NASA Earthdata account, and a free OpenAQ account to access an OpenAQ API Key. Please have these ready for Part 3 (July 22, 2026).

• Fundamentals of Remote Sensing
• Introduction to Satellite Remote Sensing for Air Quality Applications

Objectives

By the end of this training attendees will be able to:

• Match the required inputs, strengths, and limitations with different basic methods of inferring surface PM_2.5 from satellite AOD.
• Differentiate between available data products for surface PM_2.5 from NASA and Washington University in St. Louis based on their methodologies, strengths, and weaknesses.
• Use NASA Earthdata and the SatPM website to access these surface PM_2.5 data products for a location and time period of interest.
• Compare these satellite-derived PM_2.5 estimates to in situ measurements using appropriate co-location strategies and evaluation metrics.

Audience

• Researchers looking to use surface air quality estimates derived from satellites and looking to better understand the methodologies and limitations
• Air quality managers seeking to understand the potential for satellites to supplement ground-based monitors in remote areas

Course Format

• The complete course consists of three 2-hour parts, with Part 1 offered on July 8, Part 2 on July 15, and Part 3 on July 22.
• On each day, there are two opportunities to take the course (one in English, one in Spanish):
  • English Session: 10:00 a.m. to 12:00 p.m. EDT (UTC-4)
  • Spanish Session: 3:00 p.m. to 5:00 p.m. EDT (UTC-4)
• Each part includes a 30-minute live Q&A.
• Those who attend Parts 1, 2, and 3 and complete the homework by the due date will receive a certificate of attendance.

Part 1: AOD Versus PM_2.5 – When They Are and Are Not Related

• About ARSET
• Training Overview
• Part 1 Introduction
• The Pieces – Ground-based Measurements, AOD, and Chemical Transport Models (CTM)
• Why (or when) is AOD Related to PM_2.5?
• Simple Relations of AOD to PM_2.5
• Part 1 Summary
• Q&A Session

ARSET Instructor: Carl Malings

Part 2: Estimation of PM_2.5 from AOD – Methodologies and Available Datasets

• Part 1 Review
• Part 2 Introduction
• Geophysical and Hybrid Models
• SatPM_2.5 at the Atmospheric Composition Analysis Group at Washington University in St. Louis
• Machine Learning Methods
• Bias-corrected MERRA-2
• Part 2 Summary
• Q&A Session

ARSET Instructors: Carl Malings, Pawan Gupta

Guest Instructors: Aaron van Donkelaar, Junhyeon Seo

Part 3: Case Study – Comparing Available Satellite-Derived PM_2.5 Products to Ground-Based Measurements

• Part 2 Review
• Part 3 Introduction
• Overview of the Case Study Exercise
• Orientation to Google Colab
• Accessing the SatPM Dataset
• Accessing Bias-Corrected MERRA-2 CNN
• Accessing Ground-based Data
• Basic Evaluation
• Training Summary
• Q&A Session

ARSET Instructor: Carl Malings

Guest Instructor: Sebastian Diez