MIT-Data science

Satellite Data and Applications to Greenhouse Gases

Project 1: Infer methane emissions from fuel exploitation using satellite data

Background: methane is the most important greenhouse gas after CO2. The oil, natural gas, and coal industry is a major global source but the related emissions are poorly quantified. At MIT we will use advanced methods (such as Bayesian optimization, Markov Chain-Monte Carlo, machine learning, etc.) to estimate unknown causal variables (emissions, losses, etc.) from surface or satellite (such as GOSAT, TROPOMI) observed effects variables (gas concentration changes, etc.).

Project 2: Trend analysis and reconciliation of greenhouse gas observations from the global AGAGE surface network and the SCISAT satellite

Background: The MIT-led Advanced Global Atmospheric Gases Experiment (AGAGE) surface network has been measuring greenhouse gases continuously since 1978, while ACE-FTS instruments on the SCISAT have been measuring them since early 2000 from space. Here at MIT we are trying to reconcile surface and satellite data, and understand their spatiotemporal trends and climate impacts.

Semi-conduct Industry and Greenhouse Gas Emissions

Project 3: Develop a global spatial distribution of NF3 emissions from semi-conduct industry.

Background: NF3 is a potent greenhouse gas, and it is ~16000 times more potent than CO2 (in terms of GWP). Its use has grown rapidly in the manufacture of modern electronic devices (such as semiconductors, smart phone screens, etc.). We will use data mining and geocoding methods to identify point sources, estimate their emissions, derive emission factors based on production data and observations from MIT-led global network.

Project 1-3: Student should have good background on math, physics, and programing (e.g. Python). Knowledge of chemistry and climate is useful but not necessary. Students can be officially involved in MIT scientific publications if the students show innovative results in these topics.

Satellite Instrument Design for Environmental Research

Project 4: Perform Observing System Simulation Experiments (OSSEs) for a future satellite mission for constraining greenhouse gases.

Background: Observing System Simulation Experiments (OSSEs) are standard tests conducted during the design phase of a major observing program such as a satellite mission. Their purpose is to determine the ability of the proposed measurements to deliver on the scientific objectives of the program. Students will learn the general concept, structure, and models of OSSEs. Student should have basic knowledge of linear algebra, statistics, physics, and programming.

Global Chemical Transport Model and Cloud Computing

Project 5: Perform numerical simulations on Amazon AWS using a chemical transport model

Background: The global chemical transport model GEOS-Chem was first developed at Harvard University, and has been widely used worldwide, a central tool in study of chemistry-climate interactions. Cloud computing can allow GEOS-Chem users to run the model with no local hardware or systems support, and no installation. Students will be familiar with the state-of-the-art GEOS-Chem model, and get hands-on experience on numerical methods and scientific computing. Students should have basic knowledge of Linux/Unix, programing, fluid dynamics, and numerical methods. Knowledge of chemistry and climate is helpful. An Amazon AWS student account (easy to apply) is also required for this project.