Civil Engineer Turned Environmental Data Scientist Harnesses Big Environmental Data at UNESCO-IHE

Features

  • Author: Lillian Pierson, P.E.
  • Date: 08 Dec 2014
  • Copyright: Image appears courtesy of iStock Photo

At this year’s European Geosciences Union General Assembly, Dr Gerald A. Corzo presented an interesting talk on his work in hydroinformatics – or, modelling and information systems for water management. As a developer of remote modelling software and knowledge bases for UNESCO-IHE, Corzo has been using ARMA models and artificial neural networks (ANN) with a multilayer perceptron (MLP) structure to forecast and error-correct for models that simulate and forecast hydrologic discharge volumes. He’s taken account of probabilistic errors in model output and used an error corrector schema to calibrate and compare the accuracy of alternative forecasting models under different circumstances.

Dr Corzo’s integration of statistics, data visualization, programming, and environmental engineering, makes him somewhat of an environmental data science pioneer. In an interview exclusive for Statistics Views, Dr Corzo took some time to elaborate on what led him into the environmental field in the first place, and how he made the jump from traditional engineering into using a statistical approach to solve environmental water resource problems.

thumbnail image: Civil Engineer Turned Environmental Data Scientist Harnesses Big Environmental Data at UNESCO-IHE

Can you please explain why you looked outside of traditional engineering approaches, and started using statistical and data science methods to help you solve environmental problems?

Ever since I was in University, I’ve been developing algorithms for solving different types of engineering problems. In coursework, I always tried to incorporate the latest technologies to solve different types of problems, from those that arise in geotechnics, to those that are involved in even the simplest topographic applications. At the beginning, I was mostly using only computing power. Later on, however, when I started my Hydroinformatics program at UNESCO-IHE, I began using a data-driven modeling approach.
This is the way we learn machine learning in Hydroinformatics. The course at UNESCO-IHE covers machine learning, artificial intelligence, and other data-oriented methodologies to classify, group or model phenomena. The first approach was to improve forecasting by using a combination of models. Then I learned a mixture of different expert models. These ideas, trainings, and methodologies later led me to the idea of combining models to describe and forecast the hydrological process. Nowadays, I use combined models in pattern recognition and radar tracking for climate change analysis of spatial information.

If you were to advise a traditional environmental scientist or engineer on how they could make a break into environmental data science, what steps would you recommend they take? What resources might be helpful to them?

I would recommend them to start by collecting and/or organizing their data, and then thinking about what they could gain by classifying their information in new and interesting ways. They should consider successful analysis approaches that they’ve used in the past, and how those successes could be reproduced into new successes by applying a similar approach to other sets of environmental data in their collection.

It’s not so difficult to get started in using data mining to explore your data and begin generating ideas on the types of problems, limitations, and questions that the data might pose. Downloading, installing, and exploring a program like R-Studio is a good way to get going. Taking a deeper look into your data will help you extract and generalize conceptual thinking on your subject matter, and also it will help you to see groupings, complexities, and correlations of which you may not have previously been aware.

What inspired you to take a career in the environmental field? What future do you imagine for the hydroinformatics field, in particular?

My interest in the environmental field started simply as an extension of my career in engineering and physics. I was involved in the practical applications of civil engineering structures. The main link that shifted my focus from civil engineering into environmental, was hydraulics and fluid mechanics applications. (Hydraulics and fluid mechanics are considered to be both civil and environmental engineering.)

In general, most water-related processes are quite complex. In recent years there’s been a growing need for high tech solutions to water-related problems. The increasing requirement for such high tech solutions is due to the huge volumes of environmentally relevant data that are generated by satellites, telemetric machines, and even drones that are used to measure extreme weather events. Data is everywhere now. The work and new research that is being done in hydroinformatics is aimed specifically at facing challenges that are related to big environmental data.

The work and new research that is being done in hydroinformatics is aimed specifically at facing challenges that are related to big environmental data

The challenge of big data and data mining for environmental projects is the most pressing challenge that I foresee in the near future. I’m motivated to work in environmental sciences because I’m passionate about assessing whole, natural environmental systems in order to start to slowly understand how natural forces and the hydrologic cycle directly affect those systems, and thus how they affect life itself.

The challenge of big data and data mining for environmental projects is the most pressing challenge that I foresee in the near future

Related Topics

Related Publications

Related Content

Site Footer

Address:

This website is provided by John Wiley & Sons Limited, The Atrium, Southern Gate, Chichester, West Sussex PO19 8SQ (Company No: 00641132, VAT No: 376766987)

Published features on StatisticsViews.com are checked for statistical accuracy by a panel from the European Network for Business and Industrial Statistics (ENBIS)   to whom Wiley and StatisticsViews.com express their gratitude. This panel are: Ron Kenett, David Steinberg, Shirley Coleman, Irena Ograjenšek, Fabrizio Ruggeri, Rainer Göb, Philippe Castagliola, Xavier Tort-Martorell, Bart De Ketelaere, Antonio Pievatolo, Martina Vandebroek, Lance Mitchell, Gilbert Saporta, Helmut Waldl and Stelios Psarakis.