|
|
|
|
The measurement of the concentrations of different types of aerosols in the Earth’s atmosphere is crucial for modeling radiative transfer, advance allergy alerts or manage aviation circulation. This is a kind of the inverse light scattering problem, that has been approached by Machine Learning only for some very simple cases of non-spherical particles, such as spheroids.
Non-spherical particles have been widely proved to give radically different results in radiative transfer in different astrophysical scenarios, such as planetary atmospheres and cometary comae. Light scattering by this kind of particles can be modeled by several approximations, namely T-matrix for spheroids, Superposition Theorem for aggregates of spheres, Discrete-Dipole Approximation and Finite Difference Time Domain for any kind of grains. But computational limitations make it impossible to simulate light scattering by realistic irregular particles. Scattering matrices by such grains must be obtained experimentally.
The aim of this project is to develop a Machine Learning light scattering inverse problem solver that accounts for the concentration of volcanic ash, desert dust and pollen in an area of the Earth’s atmosphere based on local ground measurements. A second objective is the concept design of an apparatus that could perform such measurements and calculate the inversion. The development and patent of such apparatus would be beyond the scope of this thesis project.
The thesis project would be developed within the Light Scattering Group of the Department of Solar System at the Instituto de Astrofísica de Andalucía – CSIC (Spain), where we conduct a national project entitled “Laboratory Experiments, Observations and modeling of cometary Dust: A new Strategy (LEONIDAS)”. Studies extend from cometary tails, including exploitation of OSIRIS and Giada-Rosetta data and participation in upcoming missions such as Comet Interceptor to circumstellar disks. The student will also benefit from our collaborations in Finland and UK. Our group operates the world- class facility COsmic DUst LABoratory (CODULAB), and instrument that produces light scattering matrices by non-spherical particles, which has produced measurements of light scattering by volcanic ash, desert dust and pollen.
Requirements:
• Programming skills: Python.
• Background: Physics or Engineering.
Tasks:
• Use of available approximations for the generation of light scattering matrices of a collection of spherical, non-spherical and irregular particles matching some properties of the grains in the atmosphere (volcanic ash, desert dust and pollen). Codes are already made but their use implies a deep knowledge on light scattering and an intensive computation that will require the usage of supercomputers.
• Proposal of light scattering measurements at the CODULAB of particles relevant for the Earth’s atmosphere, as different kinds of pollen. Measurements would be carried out by those people in the group with expertise in the usage of the apparatus, so the student will not have to acquire this expertise too.
• Development of a new Machine Learning method in order to calculate the amount of each phase of aerosols in the atmosphere based on ground measurements.
• Concept design of an apparatus that could perform light scattering measurements of the aerosols in the atmosphere from the ground and calculate the light scattering inversion to give the concentration of each type of aerosol.
The student would work based at the IAA and would have access to the CODULAB. The student would receive training on light scattering and Machine Learning. The student will make stays in centers with which we collaborate, (Finland and UK) and will attend relevant astronomical schools and will participate each year in international astronomical meetings to present research results.
Group Leader
1. Title: Dr.
2. Full name: Daniel Guirado Rodríguez, Juan Carlos Gómez
3. Email: dani@iaa.es, jcgomez@iaa.es
4. Research project/ Research Group website (Url): https://www.iaa.csic.es/scattering/
5. Website description: Webpage of the Amsterdam-Granada Scattering database and
homepage of the Cosmic Dust Laboratory (CODULAB). This webpage works as an open- access repository of experimental measurements of the scattering matrix of cosmic dust analogs and of atmospheric aerosols performed at CODULAB. A description of the CODULAB apparatus and the theoretical background behind scattering matrix measurement techniques can also be found in this webpage, as well as a list of peer- reviewed publications of the group and of links to other databases and webpages of international collaborators.
Additional websites
1. Url: https://www.iaa.csic.es/en
2. Website description: Webpage of the Andalusian Institute of Astrophysics (Instituto de
Astrofísica de Andalucía, IAA). IAA is the leading Astronomy institute of the Spanish Research Council (Consejo Superior de Investigaciones Científicas, CSIC). The webpage describes research and technical activities carried out by the four Scientific Departments and the Instrumental and Technological Development Unit, including laboratory research facilities like CODULAB. It also provides information about space missions with involvement of IAA-CSIC, and the two Observatories run by the institute (Observatorio de Sierra Nevada and Observatorio de Calar Alto).
IAA is an equal opportunity institution. Applications to this program by female scientists are particularly encouraged.
Should you need any further information or assistance concerning the application, please contact the IAA at severoochoa[at]iaa.es