Education
2016 Ph.D. Earth and Planetary Sciences
Department of Earth and Planetary Sciences (EPS),
Johns Hopkins University (JHU), Baltimore, Maryland,
USA
Advisor(s): Benjamin F.
Zaitchik
Dissertation: Applications of Climate Regionalization:
Statistical Prediction & Patterns of Precipitation Variability in
Observations & Global Climate Models
2013 M.A. Earth and Planetary Sciences
Department of Earth and Planetary Sciences (EPS),
Johns Hopkins University (JHU), Baltimore, Maryland,
USA
Advisor(s): Benjamin F.
Zaitchik
2011 M.Sc. Aerospace Engineering
Department of Aerospace Engineering (AER), Cairo
University (CU), Giza, Egypt
Advisor(s): Atef O. Sherif,
Basman M. N.
Elhadidi, and Hamdy A. Kandil
Thesis: Ensemble Forecasting & Data Assimilation in
Numerical Weather Modeling for Egypt
Examining Committee: Atef O. Sherif, Jimy Dudhia, and
Mohamed E. Elraey
PDF:
M.Sc. Thesis 2011 & Evaluation
Report
2003 B.Sc. Aerospace Engineering
Department of Aerospace Engineering (AER), Cairo
University (CU), Giza, Egypt
Advisor(s): Atef O. Sherif
Graduation Project: Terrain Aerodynamics
Experience
2024 – Present Customer Contact Demand
Forecasting
Forecasting Science, Worldwide Capacity Planning (WWCP), Amazon
Customer Service (CS), Seattle, Washington, USA
Develop an automated, scalable, scientific workflow that
significantly enhances customer service operational efficiency and
decision-making, leading to optimized resource planning and cost
management.
2022 – 2024 Customer Journey
Sales, Marketing & Global Services (SMGS) Ops –
Data Platform & Infrastructure (DP&I)
Science & Econ, Amazon Web Services (AWS), Herndon, Virginia,
USA
Customer Journey enables the automation of actionable insights,
predictions, and recommendation strategies. I started with a vague
idea, just like building a successful startup, and delivered
high-quality results with long-term vision. My solution uncovered
the stages, inflection points, dwell times, insights, and drivers
of end-to-end customer journeys.
2020 – 2022 Tracking and Modeling of
COVID-19 Pandemic
Department of Earth and Planetary Sciences (EPS),
Johns Hopkins University (JHU), Baltimore, Maryland,
USA
COVID-19 Supplement to "Determinants of Enteric Infectious Disease:
a GEO Platform for Analysis and Risk Assessment" is a project to
develop an environmentally informed risk monitoring and early
warning application that will inform decision makers for
appropriate interventions and investments needed to prepare
properly for potential reemergence of the disease as policy
measures are relaxed. Machine learning and epidemic modeling are
used for generating risk maps as well as prospective tracking and
modeling of the impacts of hydrometeorological factors on COVID-19
pandemic.
2019 – 2023 Leveraging Earth Observations
for Improved Climate Projections
Department of Earth and Planetary Sciences (EPS),
Johns Hopkins University (JHU), Baltimore, Maryland,
USA
GMELT Ahead: Leveraging Earth Observations for Improved Climate
Projections in High Mountain Asia is a 3-year project to generate
high-resolution projections of future climate and hydrology that
grounded in best-available historical observations and
understanding of atmospheric processes. Different approaches are
used for downscaling, including regionalization and Convolutional
Neural Network (CNN) pattern reconstruction.
2019 – 2022 Multi-scale Prediction of Flash
Drought in the United States
Department of Earth and Planetary Sciences (EPS),
Johns Hopkins University (JHU), Baltimore, Maryland,
USA
PREEVENTS Track 2: Collaborative Research: Flash droughts: process,
prediction, and the central role of vegetation in their evolution
is a 3-year project to advance efforts to understand and forecast
flash droughts (FD) by targeting three characteristic features: (1)
land surface memory is a key component of FD, (2) evaporative
demand is a leading driver of FD onset, (3) vegetation plays a
central role in FD development through its influence on soil
moisture and turbulent heat fluxes. Deep Learning algorithms are
used for the prediction and classification of flash droughts.
2017 – 2020 Applications of Deep Learning to
S2S Prediction & Downscaling
Department of Earth and Planetary Sciences (EPS),
Johns Hopkins University (JHU), Baltimore, Maryland,
USA
Convolutional Neural Networks (CNN) are implemented and optimized
with a custom loss function (based on objective climate
regionalization) to improve dynamical forecasts at subseasonal to
seasonal (S2S) timescales. Training is performed using historical
observations and the North American Multi-Model Ensemble (NMME)
monthly forecasts.
2017 – 2020 Environmental Determinants of
Enteric Infectious Disease
Department of Earth and Planetary Sciences (EPS),
Johns Hopkins University (JHU), Baltimore, Maryland,
USA
Determinants of Enteric Infectious Disease: a GEO Platform for
Analysis and Risk Assessment is a 3-year project to develop an
environmentally informed risk monitoring and early warning
application that will inform decision makers for appropriate
interventions and investments needed to reduce enteric infectious
(EID) diseases. The project will leverage data from the MAL-ED
project funding by the Bill and Melinda Gates Foundation and bridge
this health information with environmental data from earth
observations. Very comprehensive model development to include a
wide range of predictor variables as well as targeting 8 countries
within the MAL-ED sites, generating a global picture of EID.
2017 – 2018 Advanced Seminar in Remote
Sensing
Department of Earth and Planetary Sciences (EPS),
Johns Hopkins University (JHU), Baltimore, Maryland,
USA
The fall 2017 iteration of Advanced Seminar in Remote Sensing
focused on the application, interpretation, and visualization of
Land Data Assimilation Systems (LDAS). Through lectures, exercises,
and a semester project, students learnt the theory behind LDAS, run
LDAS simulations using the NASA Land Information System (LIS), and built web
visualization applications for LDAS output using the open-source
Tethys scientific visualization platform.
2016 – 2018 The NASA Land Information System
(LIS)
Department of Earth and Planetary Sciences (EPS),
Johns Hopkins University (JHU), Baltimore, Maryland,
USA
Contributions to the development and training workshops of the NASA
Land Information System (LIS). For example, a workshop
for Egyptian applications was held at the National Authority for
Remote Sensing and Space Sciences (NARSS)
in Cairo, Egypt, in August 2017. The program included a
plenary session on August 6 and interagency technical working
sessions on August 7-10. In addition, targeted technical
preparatory activities were held in the week preceding the workshop
and follow-up project oriented activities took place in the week
after the formal workshop. A similar workshop was held at Korea
Water Resources Corporation (K-Water), Daejeon, South Korea in
January 2018.
2015 – 2016 Porting NU-WRF
to HHPC &
MARCC
Department of Earth and Planetary Sciences (EPS),
Johns Hopkins University (JHU), Baltimore, Maryland,
USA
Porting NASA-Unified Weather Research and Forecasting (NU-WRF)
model to JHU Homewood High Performance Compute Cluster (HHPC)
and to Maryland Advanced Research Computing Center (MARCC).
2015 – 2016 Climate Regionalization of
Africa
Department of Earth and Planetary Sciences (EPS),
Johns Hopkins University (JHU), Baltimore, Maryland,
USA
Regionalizing Africa based on interannual variability of
precipitation: spatial patterns of precipitation variability in
observations and global climate models (GCMs) at different times
from geological periods to historical simulations and future
climate projections.
2013 – 2015 Objective Climate
Regionalization
Department of Earth and Planetary Sciences (EPS),
Johns Hopkins University (JHU), Baltimore, Maryland,
USA
Development of an open-source R package for
Hierarchical Climate
Regionalization (HiClimR) to facilitate
the application of rigorous regionalization for climate studies. It
adds several features including multi-variate and hybrid clustering
and a new clustering method (called, regional linkage) to
hierarchical clustering (hclust function in stats
library) in R, for climate
regionalization. HiClimR is applicable to
any correlation-based clustering.
2011 – 2013 Seasonal Precipitation
Predictions
Department of Earth and Planetary Sciences (EPS),
Johns Hopkins University (JHU), Baltimore, Maryland,
USA
Application of different statistical models, including artificial
neural networks (ANN), to understand and predict seasonal rainfall
anomalies as a function of large-scale indices of surface air
temperature anomalies (SATA), sea surface temperature (SST),
surface pressure, and other variables.
2010 – 2011 Prediction of Dust/Sand
Storms
National Authority for Remote Sensing and Space Sciences
(NARSS),
Cairo, Egypt
Development of a framework for dust/sand storms prediction using
numerical weather prediction and remote sensing technology.
2010 – 2010 Porting WRF to
EUMEDGRID
Africa 4 2010 - Joint
EUMEDGRID-Support / EPIKH
School for Application Porting, Cairo, Egypt
Porting the Weather Research and Forecasting (WRF) model
to
EUMEDGRID.
2010 – 2010 High Performance Computing
(HPC)
IBM-Egypt
and National Authority for Remote Sensing and Space Sciences
(NARSS),
Cairo, Egypt
IBM AIX 5L system administration and
running Code_Saturne
Computational Fluid Dynamics (CFD) Solver on NARSS Blue-Gene/L .
2008 – 2010 Ensemble Forecasting
National Authority for Remote Sensing and Space Sciences
(NARSS),
Cairo, Egypt
Development of a preliminary ensemble forecasting system for Egypt,
which can be developed for operational use.
2008 – 2008 Estimation of Evaporative
Rates
National Authority for Remote Sensing and Space Sciences
(NARSS),
Cairo, Egypt
Evaluation of Lake Nasser water loss by evaporation using numerical
weather prediction and remote sensing technology.
2006 – 2008 Data Assimilation
National Authority for Remote Sensing and Space Sciences
(NARSS),
Cairo, Egypt
Implementation of the conventional and remotely-sensed
observational data into the numerical weather modeling system for
Egypt using Four-Dimensional Data Assimilation (FDDA).
2005 – 2006 ATOVS
Data Processing and Visualization
National Authority for Remote Sensing and Space Sciences
(NARSS),
Cairo, Egypt
Development of an automatic framework for the processing and
visualization of NOAA/ATOVS
satellite data.
2001 – 2003 Terrain Aerodynamics
Department of Aerospace Engineering (AER), Cairo
University (CU), Giza, Egypt
Generation of a surface grid for Cairo area from raster maps,
measuring flow over prototypes in a wind tunnel, and comparing the
numerical and experimental results.