Why use vector tiles for your next web map?

Esri recently released a Beta version of vector map tile technology as an alternative to the conventional raster tile cache technology that we’ve come to both love and hate over the years. With billions of web maps having been made from raster map caches, what are the advantages of switching to a vector map tile cache? Read this blog to find out what a vector tile is and why they’re an improvement to the traditional raster tile system.

As a GIS professional, you may be familiar with what a GIS vector is and how it’s used in a GIS. I’m sure you remember studying vectors in your high school algebra class. Basically, a vector is quantity that has both magnitude and direction. In GIS terms, a vector is similar in that it has a start and end point, and these points usually represent a position on Earth. Vectors can represent a single line, or the points can be combined to form a non-straight line. As well, the vertices can have a common start and end point which makes the vectors represent a polygon or a closed area.

Vectors are very common in GIS, and they are generally used as one of the basic data models for storing, finding and manipulating GIS data. On the other hand, raster data is a spatial data model that defines space as an array of equally sized cells arranged in rows and columns, and composed of a single or multiple bands. Unlike a vector structure, which stores coordinates explicitly, raster coordinates are contained in the x-y ordering of the matrix cells. 

So then, what is a raster map cache? Each screen of a raster web map is made up of hundreds of small images called tiles, which are stored individually in the web map repository called a tile cache. These tiles are 256 pixel by 256 pixel images that are pre-rendered, stored in the cache and served up and displayed for the user when the user wants to look at a particular part of the world. When the user zooms or pans, new tiles are requested, served up and rendered to the user’s display device.

When web maps were first introduced, they were amazing. Users could see any part of the world. They could zoom in or out, and the performance was really good. The reason performance has always been good is that most of the work to create the map has already been done because the basemap was pre-rendered and stored on the server. Creating a multi-scale map cache for all the world can take a long time and requires a lot of computing power to get the job done in a reasonable amount of time. I remember being ecstatic many years ago because we were able to cache all the map tiles for Canada in one week using two dedicated computers. 

Vector tile basemap (Beta) of Halifax, NS from ArcGIS Online. Note the crisp map annotation, the various street styles and the underlaid raster tile-based relief hill shade.

So, what’s wrong with raster tiles, and why is a vector tile system being developed? First and foremost, a vector tile system is similar to a raster tile system in that it is pre-rendered at all the necessary map scales. It is cut into small manageable tile size files and stored in a web map cache for quick delivery to the user. But this is where the similarity stops.

Raster tiles are fixed and offer very little manipulation other than viewing. On the other hand, vector tiles can be stylized on the fly (change the line width or colour, etc.), and the orientation of vector tiles can be changed so that areas of interest can be angled differently. Vector caches are compressed, so storage space and data transmission times are reduced. They require relatively more computing power on the client or user side, which makes them very responsive and fast. Another advantage is that vector tiles work well on any device, desktop, tablet or mobile. Check out the four major reasons for using vector tiles.

You can even make your own vector tiles using ArcGIS Pro. 

Are you ready to try out vector map tile cache technology? Nothing special is required to view vector tiles, but you’ll need a modern browser. Try a Beta version today.