TDS Special Feature

Our Shifting Global Village

Expanding cities and a shrinking countryside

TDS Editors

Published in

Towards Data Science

8 min readMar 15, 2023

Image from Unsplash, created by Louis Reed

Are you a city dweller? You’re not alone! Our global urban population has reached 56% and is expected to soar to 70% of all people by 2050.¹ With our world’s human population to reach 9.7 billion by then,² this shift in where we live will have a profound effect on how we live. We’re already seeing smart cities, ML models, and geospatial monitoring contribute to intriguing new ideas aimed at sustainability and vibrant, liveable communities.

If you work in data science, machine learning, or AI and wish to learn more about how your skills might help shape the conversation, or if you’re looking for inspiration to join this effort but aren’t sure where to start, you’re in the right place.

Drawing on TDS’s deep archive, we’ve collected 30 articles to give a well-rounded view of data science and population-related issues. We’re sharing approaches and practical solutions from geospatial analysts, architects, and authors working in specialized fields such as urban planning, supply chain optimization, and wildlife conservation.

We’ve loosely organized these articles into topics that present a good mix of global and local challenges our data scientists are tackling. Browse the titles and pick what interests you, or dig deep into a single topic. No matter how you approach this resource, you’re sure to learn something new.

When you’re ready to take your first step in writing about your own population-issue related analysis or solution, you’ll find a further list of 9 helpful walkthroughs and datasets at the end of this guide.

We encourage you to leave comments on the articles you read, and to share them — and this curriculum — as far and wide as you can.

Let’s get started!

1. Population Analytics

We begin with a birds-eye global view using population analytics and visualizations. It’s an important first step as understanding how our population is distributed, especially compared to resources, is the key to finding tailored solutions that deliver the biggest impact.

Many of these articles focus on geospatial analysis, while others offer graphical or 3D approaches. Each article offers a solid walkthrough and a good dataset to get you started.

Visualising Global Population Datasets with Python by Parvathy Krishnan (8 minutes)
High-resolution raster and vector data can provide detailed population analysis at a localized level.
Creating Beautiful Population Density Maps with Python by Adam Symington (7 minutes)
Large-scale data visualizations can be challenging! Careful selection of data and Python plotting parameters can create great population density maps.
On the grid: Estimating population density for anywhere on earth by Nick Jones (9 minutes)
Gridded datasets to zero in on local population density, a key measure in urban planning and policy.
Data Storytelling with Population Visualizations by Emily A. Halford (12 minutes)
A 3D approach to visualizing population growth over time as an effective storytelling technique.
Create Population Pyramids for Any Country with the US Census Bureau Data by Randy Runtsch (9 minutes)
Dynamic population pyramids using census data to drill down into the specifics of a population over time.

2. Urban planning

Next, we zero in on cities where population analytics take center stage in planning everything from housing and transportation to commercial services and utilities. We’re sharing articles that deal with urban planning issues, smart cities, and innovative approaches to architecture.

Computational Creativity by Karen Asmar (14 minutes)
An intriguing approach to GANs in architecture which may reveal “unthinkable architecture” for our future cities.
How to Create and Use Isolines for Profiles by Helen McKenzie (8 minutes)
The use of isolines to determine service accessibility within a catchment, an important factor in successful urban development.
Forecasting Long-Term Daily Municipal Water Demand by Blake VanBerlo (13 minutes)
An inside look at how the City of London, Canada uses ML forecasting models to predict water demand.
Energy Management Using Apparent-Time Non-Intrusive Load Monitoring by Lindo St. Angel (15 minutes)
The feasibility of non-intrusive load monitoring with the goal of a “sustainable and scalable power grid” at the consumer level is the focus of this deep learning model.
Comparative Analysis of Energy Optimization Levels Kristjan Eljand (7 minutes)
Smart energy control optimizations have the potential to significantly drop a household’s energy costs — including those associated with charging electric vehicles.
How AI Can Help Smart City Initiatives by Tirthajyoti Sarkar (8 minutes)
A deep dive into the role of AI to deal with the vast amount of data associated with smart city initiatives that help planners design everything from traffic management to water resources.
Estimating Solar Panel Output with Open-Source Data by Ang Li-Lian (9 minutes)
LiDAR can be an effective method of estimating solar panel outputs as the author of this comprehensive guide explains.

3. Sustainable agriculture

Every country faces its own unique challenges in ensuring a reliable food supply, with many depending on agricultural imports from other countries. We truly are a global village! Farming practices have already begun to shift because of climate change, and we’ll see more changes as we strive for ultra-efficient and sustainable food production. Don’t take our word for it — these authors deliver insightful articles about the future of modern agriculture.

Agronomics Optimization and Sustainability in Agriculture by Bonny Nichol’s (7 minutes)
Predicting how soil, rainfall, seeding, and fertilizer application will impact yield provides a data-driven best practice that may help farmers see higher crop yields.
Monitoring Water Scarcity (SDG 6.4) in Africa with Google Earth Engine and FAO WaPOR data by Iman Tantawy (5 minutes)
Geospatial monitoring to measure effective water usage to meet agricultural needs.
A More Accessible and Replicable Method for Satellite-Based Mapping of Hand-Harvested Crops in California by Madeline Lisaius (12 minutes)
The impact of labor shortages on hand-harvested fruit and vegetable crops using satellite imagery to find shifts in planting patterns.
Tackling the Global Food Challenge With a Data Strategy — Story of John Deere by Ekhtiar Syed (6 minutes)
Smart farming with AI, AR, and remote monitoring playing important roles may be the future standard.
Helping African Farmers Increase Their Yields Using Deep Learning by Patrick Kalkman (12 minutes)
A deep ensemble learning model for detecting plant disease from images that would allow for better crop health.
Do Neural Networks Dream of Falling Snow? by Fraser King (13 minutes)
New research into using ML to predict snowfall, a key source of regional freshwater for agriculture.

4. Our changing natural environment

As we take over more land to house and feed our growing population, our natural environment is impacted. We are part of a complex and fragile ecosystem that needs to be carefully maintained for the benefit of all. These authors present innovative management and monitoring strategies for wildlife and vegetation, as well as a close look at the impact of land-use change on atmospheric carbon levels.

Bridging the Land-Use Gap in Carbon Inventories with Satellite Imagery by Guilherme M. Iablonovski (9 minutes)
GIS techniques to capture the impact of land use change on the carbon removed from the atmosphere by urban trees, forests, and soil.
Monitor Vegetation with Google Earth Engine by Sixing Huang (10 minutes)
How satellite imagery can be used to evaluate and monitor ecosystem health.
AI Geospatial Wildfire Risk Prediction by Theo Jaquenoud (16 minutes)
A wildfire risk prediction model using raster data to minimize the hazard of raging fires and their associated costs.
The Shape of Population Dynamics by Francesco Palma (7 minutes)
Identifying and managing at-risk wildlife populations through topological data analysis.
Unveiling Fishing Activity Risk to Marine Megafauna with Geospatial Technologies by Bryan R. Vallejo (7 minutes)
Geospatial data analysis as an effective tool in protecting and conserving marine resources as expanded fishing competes with marine life.
Building a Simple AI-powered, In-the-Loop System to Manage Wildlife Camera Trap Images and Annotations by Abhay Kashyap (19 minutes)
Computer vision as a cost-effective solution to the expensive, labor-intensive task of people annotating the millions of photos captured by wildlife cameras, an important first step as wildlife natural habitats shift.

5. Transportation

Finally, we look at how people and commodities travel both on a local and global level. There’s an environmental cost to vehicles, ships, and planes that has to be considered as our population grows, but safety also becomes an issue as we see increased traffic on roadways. The authors of these articles analyze the current situation and present solutions such as bike sharing, public transportation, smart roads, and supply chain optimization.

Scrutinising Airline Efficiency by Visualising Public Aviation Data by Aine Fairbrother-Browne (10 minutes)
A comparison of airline flight data before and during COVID to identify the prevalence of “ghost” flights of less than 10% capacity, which carry a high environmental cost.
Supply Chain Optimization with Python by Samir Saci (10 minutes)
Minimizing a company’s carbon footprint while also meeting demand and reducing transportation cost is the focus of this optimization strategy.
Spatial Data Science: Network Analysis for Transportation Planning by Sutan Mufti (6 minutes)
Optimizing transportation routes and simulating the impact of new/changing routes on accessibility using a graph theory approach.
Supporting Para-transit Mobility in Africa Through Big Data Analysis by Marco Brambilla (6 minutes)
Big data may play a part in solving transport poverty by delivering reliable, affordable transportation services.
Smart Roads: How AI in Transportation Keeps Drivers Safe by Andrey Koptelov (6 minutes)
Can AI deliver smart roads that keep people safe as the number of vehicles increases alongside our population?
Producing Insights with Generalized Additive Models (GAMs) by Alvaro Peña (10 minutes)
A walkthrough of how a Generalized Additive Model can be used to promote “zero-emission transportation” by making bike sharing an easy and convenient choice.

Want to get started on your own article?

We hope the articles in this special feature have you eager to try your own analysis. If you’re unsure where or how to start, we have you covered!

First, we’re sharing two terrific population-related datasets to explore:

Next, we’ve collected a few articles from our authors who have shared datasets they’ve found or provide step-by-step guides to help you get started with a particular topic. They’re great choices to kick-start your article — and perhaps be featured in our next newsletter!

We hope you’ve already found a few articles of interest in our reading list, and that you’re inspired to tackle a topic — or two. When you’re ready to share your own data science analysis of a population-related challenge with our community, we encourage you to submit your article.

We’re excited to see your vision of the future in our shifting global village!

References

The World Bank, Urban Development, last updated on October 06, 2022.
United Nations, Global Issues: Population, accessed March 13, 2023.