#KB Geospatial Analysis — Real Estate

Dear fellow Data Scientists,

Real estate investment decisions were once dominated by intuition and historical data, but today’s firms now rely on sophisticated tools that reveal patterns beyond simple proximity to amenities. Geospatial analysis, for example, allows investors to move past broad assumptions and pinpoint specific spatial relationships that directly impact property values. This method not only identifies neighborhood dynamics and high-growth areas but also leverages predictive technologies like machine learning to forecast future trends.

Using Geographic Information Systems (GIS) and spatial modeling, investors can better predict long-term value by analyzing the spatial dynamics that influence growth. This article, written by 👋 Tiffany Le, gives a clear overview of how geospatial analysis supports smarter real estate investment decisions.

What is Geospatial Analysis?

Geospatial analysis is a beautiful art of gathering, visualizing, and manipulating GPS, satellite photography, imagery, and historical-geographical data. Geographic Information Systems (GIS), a computer system, can capture, store, and display this type of data [4]. Later in the article, we will get some hands-on practice with using GIS.

What is its role in Real Estate?

For a moment, think about it: how far away are the nearest grocery stores, pharmacies, public transportation, medical facilities, and general department stores (e.g., Target or Walmart) from your place or university? Are they located within a reasonable distance? You would ask these questions, especially if you don’t have access to an automobile.

For worried and concerned parents, the proximity and ease of accessibility to essential services and facilities around a university are appealing. Investors often prioritize investment in properties with that appeal, where geospatial analysis is handy.

Geospatial analysis can determine the optimal locations and suggestions for various tasks — identifying the best-fit in-flight entertainment based on a flight, choosing a location for a new restaurant, or even assisting disaster prevention (e.g., flood, earthquake, and other natural disaster).

Key Components of Geospatial Analysis:

There are four critical components of geospatial analysis: data collection, data processing, data visualization, and spatial modeling.

1️⃣“The Data” (Data Collection)

Geospatial data has three distinct types: point, vector, and raster. See Figure #1 for a visual example of each data type. Additionally, geospatial data is found in multiple public or private sources (e.g., satellite images, existing databases, APIs, etc.) to create a dataset.

Figure #1 shows a visual example of each data type. Points visualize point data. Irregularly shaped lines and polygons represent vector data. A grid represents raster data where each square has a different value (e.g., average temperature).

• Point data represents specific locations (e.g., schools, hospitals, and houses) where a data point corresponds to the precise geographic coordinate of a location.
• Vector data represents irregularly shaped data (e.g., boundaries of counties, states, etc.), where points, lines, and polygons represent the geographic features.
• Raster data represents data in a grid (matrix) format (e.g., temperature, earthquake risk, etc.) where each square equates to a specific value.

2️⃣Processing the Data

This step includes cleaning raw data, fixing errors, filling missing values, and combining datasets. Inconsistencies such as mismatched coordinate systems or inaccurate geotags can skew analysis results, which may need manual correction or the use of tools for alignment. Missing spatial data often requires interpolation to estimate values from nearby points. Data transformation may also mean reprojecting data to a shared map projection so all layers align for spatial analysis.

3️⃣Visualizing the Data

This step refers to creating maps to visualize the different trends, relationships, and interactions the data points can have. In GSA, data visualization is crucial since it helps us understand where data points lie in the context of others (e.g., buffers, intersections, etc.). Later on, we will discuss some specific applications of this.

4️⃣Spatial Modeling

This final step refers to analyzing spatial relationships and predicting future trends using various statistical algorithms & various methods. Some examples are geostatistics, spatial regression, and machine learning.

Key Geospatial Data Sources — Real Estate

There are various ways of accessing geospatial real estate data: public and private data sources. A great plus of geospatial analysis in real estate is that much data can be accessible for free or at a low cost. Working with public data is a great way to dip your toes in this industry. See in the below table Figure #2 a list of public and private sources that are accessible:

Figure #2: This table lists some public and private data sources.

Combining Datasets

Most likely, there isn’t one dataset that includes every feature needed for a comprehensive analysis. Situations like this are where combining datasets shines. We can incorporate features from different datasets (e.g., home prices, crime rates, population density, natural disaster risks) to learn more about the houses and their surroundings.

Geospatial Terminology

Like database management and SQL, working with geospatial data involves using tools such as intersection, join, primary keys, and so much more! A valuable and accessible tool is creating buffers. This tool draws a circle around a data point with a specified radius. With this tool, we can determine if a house is within a certain proximity of a service (e.g., public service, grocery store, restaurant, etc.).

For this case study, 0.5-mile buffers [8] were created around each school, and 5-mile buffers were created around each hospital.

Geospatial Analysis with ArcGIS

Developed by Esri (Environmental Systems Research Institute), ArcGIS is a geographic information system (GIS) that provides datasets and analysis tools for creating, analyzing, and managing spatial data [7][11]. ArcGIS is a powerful tool for visualizing and interpreting 2D or 3D spatial data [11]. Good News! ArcGIS Online is free; many universities offer access to ArcGIS Pro for more advanced use.

• Chapman students, access your free ArcGIS account here.
• Check out a similar map designed for this case study here.

Geospatial Analysis with Python

Python is a versatile language capable of handling tasks such as writing basic conditional statements, performing machine learning, and completing geospatial analysis. Python packages such as GeoPandas and GeoPy make geospatial analysis easier and more accessible [9][10].

A Competitive Edge for Investors?

Investors can use geospatial analysis to gain a competitive edge. Besides analyzing the surrounding data of a property, geospatial analysis is used to predict the future growth of a community and its surroundings [5][6]. Will there be a shopping plaza built nearby? Will a park be built here? What’s the risk of a future natural disaster occurring? These are some questions that geospatial analysis assists in predicting.

Case Study: Evaluating Real Estate Investment Opportunities in a Metropolitan Area (King’s County, WA)

Figure #3: This figure is a geospatial analysis conducted on King County, Washington homes. A green point represents a house. Blue points are houses located within a 0.5-mile radius of a public school. A red point is a house within a 5-mile radius of a hospital. Finally, a yellow star represents houses within a 0.5-mile radius of a school and a 5-mile radius of a hospital.

This case study focuses on analyzing the location of houses in King County and their proximity to local public schools and hospitals. The data was sourced from ArcGIS’s Living Atlas and Kaggle (King’s County) and analyzed using ArcGIS Online/Pro and Python [7].

Conclusion

Geospatial analysis offers real estate investors valuable insights by assessing factors like infrastructure proximity, neighborhood trends, and environmental risks. It allows for a more accurate evaluation of potential investment areas by combining data from various sources, including property values, crime statistics, and zoning changes. This helps investors pinpoint promising locations while managing risks tied to market shifts and urban development.

Thank you for reading, and happy programming!

References

• [1] Yang, Yang, et al. “Embracing geospatial analytical technologies in tourism studies.” Information Technology & Tourism 25.2 (2023): 137–150. APA
• [2] Gale, Harrison, and Shouraseni Sen Roy. “Optimization of United States residential real estate investment through geospatial analysis and market timing.” Applied spatial analysis and policy 16.1 (2023): 315–328.
• [3] Chen, Hua, et al. “Dynamics and driving mechanism of real estate in China’s small cities: A case study of Gansu province.” Buildings 12.10 (2022): 1512.
• [4] https://education.nationalgeographic.org/resource/geographic-information-system-gis/
• [5] Long, Ying, and Zhenjiang Shen. Geospatial analysis to support urban planning in Beijing. Berlin/Heidelberg, Germany: Springer International Publishing, 2015.
• [6] Altuwaijri, Hamad Ahmed, et al. “Predicting urban growth of Arriyadh city, capital of the Kingdom of Saudi Arabia, using Markov cellular automata in TerrSet geospatial system.” Arabian Journal of Geosciences 12 (2019): 1–15.
• [7] Aydin, Orhun. “Build House-Valuation Models with Machine Learning.” Learn ArcGIS, learn.arcgis.com/en/projects/build-house-valuation-models-with-machine-learning/. Accessed 29 Sept. 2024.
• [8] Lam, Tiffany. “Molly Green.” Safe Routes Partnership, 22 May 2018, www.saferoutespartnership.org/blog/too-far-walk#:~:text=Most%20Safe%20Routes%20to%20School,acceptable%20distance%20for%20high%20schoolers.
• [9] Geopandas. “Geopandas/Geopandas: Python Tools for Geographic Data.” GitHub, github.com/geopandas/geopandas. Accessed 29 Sept. 2024.
• [10] GeoPy “Welcome to GeoPy’s Documentation!” Welcome to GeoPy’s Documentation! — GeoPy 2.4.1 Documentation, geopy.readthedocs.io/en/stable/. Accessed 29 Sept. 2024.
• [11] “An Introduction to Arcgis Online.” An Introduction to ArcGIS Online-ArcGIS Online Help | Documentation, doc.arcgis.com/en/arcgis-online/get-started/what-is-agol.htm. Accessed 29 Sept. 2024.