Internship Opportunities

Algorithm development for satellite retrievals of oil slicks & other substances

The retrieval of oceanic properties from satellites is a very important component of climate research. In preparation for the NASA Plankton, Aerosol, ocean Ecosystem mission (to be launched in 2023), we are developing an advanced retrieval scheme that, in addition to measurements of intensity, exploits the polarization state of the light measured by its state-of-the-art optical sensors. The retrieval scheme belongs to the class of “inverse methods”, which can be applied as solvers to the widest class of problems and have the advantage of rigorously determining the uncertainties associated with each retrieved parameter. In our case, the Python LMFIT “inversion wrapper” drives a “forward” radiative-transfer engine (written in Fortran) and will enable the retrieval of parameters descriptive of the ocean surface like its refractive index, with the primary application of detecting oil slicks or biogenic films. Through the interaction with the GISS RSP group, the intern will have the chance to be exposed to several aspects of remote sensing for climate research, from the preparation for airborne and spaceborne campaigns to their execution and subsequent data analysis.

Code development remote sensing of snow properties

The retrieval of snow properties and their evolution in polar regions is a very important component of climate research. We are in the process of developing a new retrieval scheme that exploits the polarization state of the light measured by satellite sensors (POLDER), in addition to measurements of intensity only (like those of MODIS). Such a retrieval scheme is composed of a “forward” radiative transfer engine (written in Fortran), driven by an “inversion” wrapper available as part of a Python package. Inverse methods can be applied as solvers to the widest class of problems and have the advantage of adding a detailed error budget estimate of the state parameters to the retrieval of their values. In this case it will enable the retrieval of parameters descriptive of the snowpack like grain shape and size, the concentration of light-absorbing impurities, but also the simultaneous determination of the properties of aerosols that might be present in the scene above the snowpack.

Fire in the Climate System

The project will use a climate model and observational datasets to understand and simulate processes that determine how wildfires impact climate and air pollution. The goal is to advance the understanding of how anthropogenic and natural pollutant emissions influence atmospheric chemistry, climate, and air pollution. Data from the NASA GISS climate model will be analyzed for aerosol and gas concentrations and be compared to aircraft and satellite measurements. We will investigate science questions that either address climate change or air pollution.

Merging and analysis of multi-sensor imagery over polar regions

Advanced satellite retrievals of snow properties benefit from the synergistic exploitation of data originating from multiple sensors. For this reason, such data needs first and foremost to be co-located and merged into custom files for practicality of use when input to the retrieval algorithms. Continuing the work performed by previous interns, we will exploit available processing tools to co-locate several-years' worth of datapixels from the MODIS, POLDER, and CALIPSO sensors and run statistics of interest on pixel-based properties. Ideal candidates for this project are students with strong interdisciplinary skills, including experience with the analysis of geophysical datasets and their visualization, but also well versed in code development. High proficiency in Python is a requirement, and knowledge of system architecture concepts is considered an advantage since the batch processing of large amounts of data requires to be optimized for speed.

Parameterization of phytoplankton absorption for NASA/PACE retrievals

The Plankton, Aerosol, Cloud, ocean Ecosystem (PACE) mission is a NASA Earth-observing satellite mission that is scheduled for launch in 2023. The NASA/PACE spacecraft will carry three state-of-the-art instruments to monitor changes in oceanic and atmospheric particulates. Two of these instruments, the Ocean Color Instrument (OCI) and the Spectro-Polarimeter for Planetary Exploration one (SPEXone) will take ultraviolet (UV) pictures of the Earth. This is the first time that NASA will make such UV pictures to study changes in the plankton population, offering new opportunities to study how our oceans are changing on a global scale, which is both exciting and challenging. A proper exploitation of the UV data collected by the NASA/PACE sensors requires models that simulate the sensitivity of the “UV color” of the ocean to the particulates suspended in the seawater. The purpose of this project is to help create such models by providing parameterizations of phytoplankton absorption spectra. To this end, we are looking for an intern who will be tasked with using statistical models to help parameterizing an existing dataset of 700+measurements for phytoplankton absorption spectra.

RSP Data Management, App Development, and Code Conversion

NASA GISS Airborne Research Scanning Polarimeter (RSP) is often flown in field deployments. It remotely collects data to measure aerosol and cloud properties. During a field deployment RSP needs to be monitored to make sure it is healthy and collecting data properly. In addition, we can do real time retrievals so our team can contribute to the discussion of interesting scenes we observe so that perhaps we can look into it deeper. Currently, the data is processed using code written in IDL (Interactive Data Language). IDL is not commonly known language and also it is not easily portable as it requires license which most people do not have. We would like to convert this code to Python. This way the code becomes more usable and shareable. Other programming codes may need to be converted to Python as well. After the flight the data is placed on GISS web site. The site needs to make it easy for others to select relevant data of their interest. It shows ground tracks and filters data given criteria of interest. It also needs to display the data (pseudo image and plots) for a quick analysis. The intern may work on the RSP website and app.