NASA EarthRISE and Remote Sensing Interview Preparation

Geospatial, remote sensing, and NASA-specific vocabulary for intern interview preparation.

EarthRISE Academy

A JPL program that allows students and early-career professionals to contribute to NASA Earth science initiatives by translating satellite data for environmental and public policy applications.

Aerosol Optical Depth (AOD)

A measure used to analyze atmospheric content, specifically used in the Palisade Fire project to track smoke plume evolution.

LIMS

Laboratory Information Management System; used at Eurofins for managing high-throughput environmental sample tracking and data validation.

Finite Element Analysis (FEA)

A computational method used to evaluate the structural integrity and durability of components under stress, applied in the Anteater Formula Racing project.

ARSET

Applied Remote Sensing Training; a NASA program providing training resources and open-access workflows for systematic handling of satellite datasets.

SERVIR

A 'demand-driven' global model connecting NASA space-based observations with local decision-makers to tackle issues like water security and land use changes.

TOPS

Transform to Open Science; a NASA initiative focused on making big cloud-native data accessible for local community action.

NISAR

NASA-ISRO Synthetic Aperture Radar; a mission measuring Earth's surface changes down to fractions of an inch, generating massive volumes of radar data.

SWOT

Surface Water and Ocean Topography; a mission that surveys nearly all water on Earth's surface for resource management and disaster mitigation.

PACE

Plankton, Aerosol, Cloud, ocean Ecosystem; a mission providing hyperspectral imaging of ocean ecology and advanced aerosol tracking for air quality.

Coordinate Reference Systems (CRS)

Standardized spatial frameworks that must be aligned, often using $gdal.Warp()$, before performing geospatial calculations.

Nearest-neighbor Resampling

A resampling method used for discrete categorical data types such as land-cover classification.

QA/QC Bands

Quality control tools or bitmasks used to screen out clouds, shadows, or sensor errors by converting bad pixels into NaN values.

NASA

National Aeronautics and Space Administration; its founding charter objective is the expansion of human knowledge of phenomena in the atmosphere and space.

What drew me to EarthRISE program @ JPL

• embodies NASA's TOPS (Transform to Open Science) initiative.

• JPL’s unique position (combining CalTech's academic innovation with NASA's Earth observation fleet like SWOT and NISAR) creates the perfect environment to take big cloud-native data and make it truly accessible for local community action.

Factor of Safety

An engineering design value used to ensure safety; in the racing project, it was balanced against material thickness to avoid overengineering.

GIOVANNI UV Aerosol Index

A NASA data source used to confirm the presence of absorbing aerosols like black carbon in smoke plumes.

Resume/Telling them about myself + Projects

• B.S. Earth System Science (Graduate from UC Irvine)

  • Specialized in Oceanography

  • Minor in Materials Science and Engineering

• Background: Intersection of environmental domain science & technical problem-solving

• Computational Side: automated data pipelines in Python & qGIS ~ process multisensor NASA satellite datasets to analyze Aerosol Optical Depth and land cover changes for the Palisades Fire Project

• Industrial Side: Managed high-throughput data validation & LIMS pipelines under strict regulatory standards @ Eurofins + Worked with rapid prototyping and engineering design tools like SolidWorks & Linux-based systems @ AFR

Why JPL ( why work for NASA EarthRISE Developers Academy)?

• JPL has a mission to make big Earth observation data truly accessible

  • This allows students and early-career professionals to contribute to real NASA Earth science and open-science initiatives

  • Provide opportunity to combine Earth science background w/software & data skills ~NASA’s observations more usable for researchers, decision-makers & communities

Resolved conflict/technical roadblock (Palisade Fire Project)

• S: Palisade Fire Project~ wanted to analyze the vertical plume movement of fires; confirm whether pollutants were being injected to atmosphere/affecting neighboring areas such as Malibu and Santa Monica Mountaisn

• T: Original datasets for tracking smoke transport weren’t available [3D wind field data & LiDAR observations for vertical plume movement not available for 2025 timeframe ~ couldn’t directly measure vertical plume movement]

• A: Redesigned workflow to integrate multiple NASA datasets (MODIS thermal anomalies, surface reflectance products, UV Aerosol Index) & refined spatial analysis w/constraints (exact fire perimeter) ; inputted Python scripts to segment data into baseline, ignition & recovery periods

• R: Able to indirectly validate atmospheric transport & identify synchronized inc. in surf. refl. & aerosol index values ~ smoke descended —> planetary boundary layer

  • communities downwind were being exposed to wildfire related pollutants despite the absence of direct meteorological obsv.

Communicate complex datasets/remote sensing to local/community partners [Wildfire Smoke Impact]

• Avoid overcomplicating science & focus on what data means for their decision-making

  • Analogy: Satellites are like high-powered digital cameras in space

    • Normal Conditions: Satellite ~clear image

    • Abnormal Conditions: Satellite ~ image changes in measurable ways

      • Can compare changes to normal conditionals (~where smoke is concentrated and how its moving)

  • Actionable Impact: Data correlates to day-to-day operations; Rather then providing raw raster files/complex array outputs —> translate satellite readings to simple color coded maps

  • Bottom Line: Red = heavy ground-level particulate pollution (air quality hazardous) | Green = clear and clean conditions

    • Indicate which neighborhoods need emergency air filtration units or need to be closed to protect the public

    • They need to know what the satellite sees, what it means for their community, and what action they should take next."

Ex. of complex concepts to someone not familiar with the subject matter + Cross Functional Team [AFR]

• S: AFR included working with mechanical, electrical, and systems engineering students to prep a high-performance vehicle for comp; designed head restraint and harness system

• T: Evaluated safety requirements and load distributions that many team members outside the HI group were unfamiliar with

• A: Instead of just presenting the engineering calculations alone for design review, I translated the requirements of my design into visual diagrams —> how the harness system would behave via collision ~component durability & structural integrity under stress (Design decisions ←→ Driver safety & Formula SAE rulebook)

• R: Framed around safety outcomes > technical equations; team able to understand reasoning behind design —> implementation

Ex. of different perspectives & handling of situation [AFR]

• S: HI Engineer for AFR~ most optimal material to weld onto chassis for head restraint brackets (would hold the cushion drivers rest their head on)

• T: Rulebook*** + debate between me and one of my colleagues on which type of steel would be most optimal; advised current material is too thick to weld despite a large FOS (thought was good for driver but Engineering R.O.T is to push limits) reduction of thickness —> lighter car

• A: Decided to consider their viewpoint & conduct FEA w/varying thickness [find thinnest material w/good FOS]; mock setup replicated conditions highlighted in rulebook

• R: Data from FEA proved a thinner material was indeed more optimal & cost-effective to implement; strong engineering decisions come from combining perspectives & validating assumptions w/data (data>intuition)

Encountered & Resolved Anomaly? (Eurofins)

• S: Lab tech @ Eurofins; managed high-throughput environmental sample tracking & data entry in LIMS database (complied w/strict EPA & Title 22 stands.)

• T: Running routine QA/QC ver. on large batch of water quality metrics ~ discrepancy in historical baseline values & newly flagged analytical outputs

• A: Didn’t pass data downstream but audited the data trail; cross-ref. phsy. log sheets w/electronic entries in LIMS ~id. calibration offset in analytical instrument & reported issue to my manager

• R: Flagged affected samples for re-testing under standardized protocols; prevented potentially inaccurate compliance decisions from being made using flawed data & upheld strict EHS/EPA quality standards

Why they should select me over other qualified candidates?

• Bring combo of strengths that align w/EarthRISE’s goals + ESS background give understanding of env. ?’s behind the data; experience w/Python, qGIS & NASA satellite datasets allow me to work w/technical side of remote sensing workflows

• Strong commitments to data quality thru work @ Eurofins & can collaborate across disciplines thru projects like AFR; also enjoy translating complex technical info. to something accessible/actionable for diff. audiences (key to EarthRISE’s mission)

• Genuinely excited to learn; despite being early in my career, I’m someone who is curious, adaptable & willing o put in the work to contribute to the team; grateful for opportunity to bring that mindset to EarthRISE + cont. to grow as a developer and Earth scientist

Best Practices for remote environment

1. Respect your peers' time and remote situations

2. Able to work independently and stay organized

3. Maintain communication with teammates

Greatest Weakness

1. giving negative feedback

2. important but I worry about offending my teammates or making them feel like the work they did and the goals they accomplished went unnoticed

3. communication is key, constructive criticism is really valuable for growth, so I have to push myself out of my comfort zone

4. in the end we're all on the same team and I have to keep in mind that by providing that kind of feedback, I'm making the team stronger

5. something I think I can continue to improve on in DEVELOP, esp. since the teams are small and tight-knit and rely on strong communication

Raster vs Vector Data

• Raster: grid of pixels/cells ~ satellite imagery, temp. AOD & elevation

• Vector: points, lines, polygons ~ roads, fire perimeters, boundaries

  • EX: Palisade Fire Imagery = Raster | Fire Perimeter Shapefile = Vector

Coordinate Reference System (CRS)

• how locations on Earth are represented, diff. datasets use different CRS

  • CRS must be standardized b4 analysis to prevent layers from being spatially misaligned

Datasets w/different projections?

• check CRS of all datasets first, reproject to a common CRS, verify alignment visually in qGIS, use appropriate resampling methods, & validate b4 performing calculations

Missing/Corrupt Pixels

• identify source (cloud or shadows or sensor issues), use QA/QC bands/masks, convert NoData values to NaN, exclude invalid pixels from calculations, & consider alt. datasets if coverage is poor

Validate Remote Sensing Results

• compare multiple datasets

• check consistency w/known events

• cross-reference literature/ground observations

• look for expected spatial & temporal patterns

• verify outputs visually & statistically

Why multiple satellite datasets?

• no single sensor captures everything

• diff. sensors measure diff. phenomena

• improve confidence in conclusions

  • fill data gaps (ex: MODIS + UV Aerosol Index in the Palisades project)

Data quality is just as important as the analysis itself (verification is key)

Python libraries used before

• NumPy → numerical calculations

• Pandas → tabular data

• Matplotlib → visualization

• Rasterio → raster processing

• GeoPandas → vector data

• Jupyter Notebooks → workflow development

Automate GIS workflow

• prototype manually in qGIS & identify repetitive steps

• build python functions for each step

• loop thru datasets (automatically)

• generate reproducible outputs

Process large geospatial dataset

• avoid loading everything into memory

• clip to ROI first

• process data in chunks

• use cloud resources when available & work only w/necessary variables

Unfamiliar datasets

• read metadata/documentation

• understand variables & units

• visualize the data

• check quality flags

• start w/small test analyses

I focus on understanding the data before jumping into analysis