What does Data Resolution have to do with Data Accuracy in GIS?

The strength of GIS is in its data. Inaccurate or “bad” data will create weak analyses and assumptions. Data resolution and accuracy are key factors of data reliability, and they are related. High-resolution data is typically better able to represent the earth’s surface, so it is considered to be more accurate as it provides the ability to be more precise. It’s easier to precisely map the things you can see more clearly. There are many other ways that error can be introduced to a GIS project, including outdated data sources, errors in data entry, inaccurate measurements, and inherent limitations of the data collection method. This blog will specifically focus on data resolution and accuracy and how they relate. 

What are Resolution and Accuracy?

Spatial Resolution is the smallest difference between adjacent positions that can be recorded. In rasters, it’s measured with cell size as pixel height and width. GIS files measure location data in ground units (meters, feet, centimeters, etc), which translates to resolution. It’s simplest to describe this with rasters, but the same principles of accuracy and resolution apply to vector data as well. Other interpretations of resolution exist (pixels per inch, overall pixels in an image, how many megapixels), but they aren’t as important in GIS data.

Accuracy is how well the data reflects the earth it’s mapping. We can see it when an image has low resolution: objects are blurry and difficult to discern. In GIS, this can happen when pixels are larger than the details in the object that it’s depicting. Each pixel only has one value, so all of the information in that area must be averaged down. 

Resolution Affects Precision, which Affects Accuracy 

Let’s compare the high and low-resolution imagery in the image set below. It’s almost impossible to see the exact corners of the buildings in the low-resolution image, which has a spatial resolution (cell size) of 10ft. Digitizing features from the low-resolution data would involve about 10ft of guesswork. Higher-resolution map data allows users to precisely see, measure, and digitize features. 

In general, you want to have the data resolution that’s appropriate for the geographic phenomena you’re studying. The 10ft resolution data work for measuring city bounds, or the extent of vegetated area, but not for identifying buildings that are less than 10ft large. 

Downsampling Doesn’t Improve Data Accuracy 

Having high resolution doesn’t automatically make a dataset more accurate. Downsampling happens when high-resolution data is created from low-resolution data without additional information. It only creates more, smaller cells. This reasoning also applies to analysis methods. For example, if you have 30m data, an analysis done at a 10ft scale isn’t necessarily accurate to the new scale. This is because each 30mx30m sq area is represented by a single pixel with a single elevation value. All real-world elevation values that exist within the pixel are averaged into and represented by that single value, erasing local terrain fluctuations. Any analysis done with this data below a 30m resolution will have a high margin of error because it doesn’t reflect real-world (in-situ) values. 

While downsampling can be alright as long as you have the original resolution in mind, it could easily mislead a reader of the report it’s used in or a future user of the dataset. Depending on the context and use case, there’s a chance that using high-resolution data to display results on a low-resolution analysis can mislead readers. If you’re curious about the ethics of map making, read more in the book How to Lie with Maps by Mark Monmonier.

Increasing the resolution in Global Mapper’s analysis tools will not increase the accuracy of the data, but it will create more, smaller, identical cells. For higher resolution and more accurate data, you’ll need to find other datasets. Global Mapper’s Online Data tool can be useful in these instances. Data sources can be streamed directly into your workspace, or links to eternal sources automatically panned to your study area.   

Try your data Global Mapper by downloading a free, 14-day trial today!