Lana Khalifa portrait

Lana Khalifa

Environmental Engineer | Machine Learning

Portfolio Overview

Education BSc MSc Work Experience Children

Education

2021-2024

M.Sc. in Environmental Engineering

Technion - Israel Institute of Technology

Thesis: Deep Learning for Flood Modeling: Generalizing Across Terrain and Flood Event Variations (Technion Environmatics Lab, TechEL).

2016-2021

B.Sc. in Environmental Engineering

Technion - Israel Institute of Technology

Final Project: Abatement Cost Calculation of Biomass-Based Fuels through Monte Carlo Simulation (Prospective and Early-Stage LCA Lab, PSELL).

B.Sc. in Environmental Engineering

During my studies at Technion, I pursued sea-and-atmosphere electives and developed strong skills in environmental modeling, numerical methods and coding.

Academic Distinctions

  • Awarded a first-year merit scholarship for an outstanding entrance score.
  • Dean's Honor List (Spring 2018/19).
  • President's Honor List (Winter 2017/18).
Record of Excellence

Research (Student Position)

  • Co-authored a life cycle assessment study on plastic biodegradation by larvae at the PESLL lab, under the supervision of Prof. Sabrina Spatari.
First page of published article on plastic biodegradation by larvae

DOI: 10.1016/j.scitotenv.2020.139521

Final Project

  • Topic: Abatement Cost Calculation of Biomass-Based Fuels through Monte Carlo Simulation.
  • Awarded the Moshe Shnabel Scholarship for Creativity and Excellence in Civil and Environmental Engineering.
Abatement cost distribution for biomass-based fuels

Abatement-cost distributions from Monte Carlo analysis across biomass pathways. The unit is $/ton CO2, meaning the cost to avoid one ton of CO2 emissions. Pathways differ by feedstock and by conversion route.

M.Sc. in Environmental Engineering

github manuscript

Project concept presented in poster form at Faculty Research Day (March 2024).

(before implementation)

Poster presenting the project idea before implementation

Data acquisition for flood simulations

Terrain selection from high-quality DEM and dry ecoregions overlap
Available high-quality 3DEP DEM coverage
Dry ecoregions (hot, dry summers)
Elevations measured in summer

Spatial overlap used to identify terrain candidates for flood simulations.

Example of terrain tiles clipped from selected overlap areas
Overlap candidate region
Dry ecoregions
Intermittent rivers
Clipped terrain tiles used for simulations

Example: terrain rectangles clipped from the overlap of high-quality, summer-measured elevations, dry ecoregions, and intermittent rivers to define simulation domains.

Overall number of terrains: 210

simulations' setup (210 in total)

Overlayed simulation grid with sub-grid bathymetry detail

Overlayed grid (one example out of 210).

Simulation inputs including terrain data and hydrograph boundary conditions

Water inlets and example hydrographs.

RUN!

train and test

Training and validation loss curves used to compare deep learning architectures Model predictions compared with targets and error maps

preprocessing: from simulations to patches

Workflow showing transformation from simulation outputs into patch-based datasets over time

Producing and augmenting patches from one simulation result.

Input and output of one sample

Input and output of each sample.

Deep learning architectures: benchmarking and custom design

Examples

Simplified UNet
Simplified UNet architecture diagram
Encoder-Decoder with Self-Attention
Encoder-decoder architecture with self-attention
Custom Architecture
Custom deep learning architecture diagram

closure model: from patches to full domain

Closure model patch traversal sequences used to iteratively fill the full domain from patch predictions Example of closure model applied to an unseen simulation timestep

Left: A non-DL iterative closure model. Patch locations evolve at each iteration, updating next-step boundaries until the full-domain solution stabilizes. Right: Closure applied on an unseen simulation.

Work Experience

2025-2026

Senior Coordinator for Urban Flooding

The Israeli Water Authority

  • Left the role to pursue a career transition toward early childhood education.
2024-2025

Research Assistant

Technion Research & Development Foundation

  • Prepared calculations and figures for a scientific study based on an expanded version of my B.Sc. final project.
2022-2024

Teacher Assistant

Technion - Israel Institute of Technology

  • Assisted in teaching undergraduate courses: Introduction to Numerical Methods, Statistics, and Fundamentals of Fluid Mechanics.
2019-2021

Student Position

Technion Research & Development Foundation

  • Co-authored a research article on plastic biodegradation by larvae using life cycle assessment.

Children

Preschool Search

  • Currently visiting kindergartens to find a Montessori preschool assistant role.

Research Interests

  • Developing AI methods to study learning behaviors in early childhood settings (ages 3-6), supporting both individualized learning trajectories and understanding of the overall classroom state.
  • I am especially interested in multi-modal ML: computer vision + audio.

AI-Assisted Workshop Ideas

  • Developing workshop ideas, utilizing language models.

AI-generated examples:

AI-generated workshop concept 1 AI-generated workshop concept 2 AI-generated workshop concept 3

Neighborhood Implementation

  • Implementing the ideas with children in my neighborhood.
Children implementing an activity concept Children collaborating on a workshop activity