Combining Historical Maps
over the Internet

University of Technology Eindhoven, 17 nov 2009

http://mapserver.sara.nl/conferences/2009/eindhoven_tu/

Jan Hartmann
Department of Geography
University of Amsterdam
j.l.h.hartmann@uva.nl

(NB: Click the pictures to see the slides as shown on the conference in a separate browser window; some of them get very grainy when reduced)

Ladies and Gentlemen

First, let me introduce myself. I am Jan Hartmann, and I work at the Department of Geography of the University of Amsterdam as GIS-technician. My specialisation is Internet Cartography, and I am here to prepare a course Jos Bosman and I are going to give at the beginning of next year about architecture, maps and history in the Rhine-Meuse area. This presentation is a foretaste of the mapping part of that course. It's all a bit experimental, so I won't give you a highly polished Powerpoint presentation, but just an impression of my ideas about the possibilities of Internet mapping in an architectural context. However, I wrote the lecture out, and you can read it at the this URL The pictures I am showing you today are shown to the right of the text, and can be a bit grainy because of the reduction. Just click on the picture, and you get the full-sized illustration in a new window.

Then, I will sometimes ask you if you know a certain concept, computer term or idea. That is not to test how clever or stupid you are, but for me to get a feeling of what sort of knowledge I can expect from students here about cartography, computers and history. I come from a department of Social Geography in the Social Sciences, which is quite different from an Architectural Department at a Technical University. Also, the technology I am going to speak of, Open Source software, is not as widespread in University Departments as commercial software, and I would like to get some feeling about your knowledge about that too. So feel free to interrupt with questions or observations, as I will learn from them too.

I start with this picture, in the first place of course, because it is a very beautiful picture, but also because it is related to the contral technology I am going to speak of: Open Source Mapping. This picture was taken at the first MapServer conference in Minneapolis in 2003, when it was just a hacker's meeting. Nowadays, MapServer is a very sofisticated program (here you see its home page), with of course a map, showing the location of the 2010 MapServer conference, Barcelona. The thing about this map, of course, is that it is interactive: you can make it zoom and scroll. Of course you have all seen something like this: web pages with maps in which you can move around a bit, e.g. route planners and driving directions. They exist for quite some time now, but the big revolution in that technology has come with Google Maps in 2005. There are two things in Google Maps that made it really revolutionary:

1. It showed very detailed maps of the whole world, not only street maps, but also satellite views. Most of this material was acquired, but some of it was produced by Google itself, e.g. the satellite maps of Holland and Germany. Here you see the three-country-boundary between Holland, Belgium and Germany in Southern Limburg, and you can see very clearly the color differences between the countries. It was (and is) exceptional that such large scale, detailed maps are available without cost to the general public.
2. Google maed it possible, and this is really revolutionary, to use these maps in your own web application. Technically, it made available the Google Maps API, the "Application Programming Interface" for its map. Quote:
   "The Google Maps API lets you embed Google Maps in your own web pages with JavaScript. The API provides a number of utilities for manipulating maps (just like on the http://maps.google.com web page) and adding content to the map through a variety of services"
   An Api is just a set of programming instructions, in this case written for the programming language "Javascript", the language that is used to make web pages interactive. I guess you always wanted to know what an API looks like, well here you see the start of the reference page, with the basic objects: map, marker, polyline and polygon, and here you see the start of the "Marker" object.

Essentially, Google Maps is not just a mapping application, but a primitive Geograpical Information System. Geographical Information Systems are very complex and expensive programs, that mapk it possible to combine maps. The revolutionary thing about Google is that it made a small subset of the functionality freely available, together with a very detailed base map of the whole world. To give an indication of how revolutionary this is: a full fledged GIS costs round 10.000 Euro, a detailed topographical map of the Netherlands about the same, and you are not allowed to put the results on a website visible by the general public.

What I am going to talk about this afternoon is how these maps are made, and especially how you can make them yourself: how can you put your own architectural maps and drawings on the Internet, zoomable and pannable, and how can you combine them with Google Maps and the other freely available Internet maps. As I said, I am going to use Open Source technology for this, webserver technology with MapServer, database technology with PostGIS, and browser interaction with OpenLayers. What you see here is a very simple OpenLayers application, and please note the power of Open Source technology, combining maps from different commercial providers

There are several ways of combining putting your own map data on the Internet and combine them with Google maps. THe most simple one, and the one you have all seen, is just putting markers on Google maps. One of the first, and a very notorious one, is the Chicago criminal map. This version is already a bit more sophisticated, as it also shows street views of the locations selected. There are lots of political and ethical issues with this sort of map, but we won't go into them over here. I will say, however, that in your discipline, architecture and planning, there are lots of political issues concerning maps, especially their visibility to the public: we, the general public, are not able to judge assess or evaluate the way our environment is modelled by the state without detailed base maps. As I said, these are as a rule *not* available. In my opinion, one of the causes of the large scale "vastgoedfraude" you are reading so much about nowadays, is that we are not able to verify the maps on which these projects are based.

We call this sort of simple cartography on top of Google Maps a "Mashup" You see here the formal definition, essentially combining several digital media on one page. Looks great, but informally I have to revert to Dutch. According to my dictionary, the word Mash means something like "rotzooi". Nice to see of course, and to advertise with, but competely unfit for application that require precision, and architecture is certainly one of the most precise ones. For all that, they can be great fun, and people can become quite creative people by playing with maps, not to say manic. Even creating imaginary maps. This one is very nice: a mashup of ugly buildings in Belgium.

Seriously, when you look at mashups, they are almost exclusively maps of points. If we want to do some serious architectural work, we need to make precise drawings of lines and surfaces, and even of three-dimensional solids. I won't go into 3D ( here an example from London, and here from Amsterdam, here zoomed in. Here of course a bird's view of Eindhoven, with here the church and here the railway station, made by 3dClear I don't know who makes these things, but essentially their are easy to produce with Sketchup and they are almost almost as easy to place in Google Earth. The same problem here as with simple mashups: the technology is simple, that makes it very fast and creative, but if you want to work in a real precise manner, you need to put a bit more effort in the technology. I won't discuss these 3D-aspects here, but I will explain how you can prepare the ground-drawing to be compatible with Google Maps (and Google Earth, the principle is the same). Filling in the the third dimension is then a matter of drawing with Sketchup.

You all know how to make drawings and maps with stand-alone computer-applications like Autocad, but how can we get these products on the Internet, and combine them with Google Maps and other Internet maps. The simplest way is like this: just click to form points, lines and polygons, and the items are pinned to their *geographical* location. This immediately introduces the two most important fundamentals of computer mapping: All items are either points, lines or polygons, and all items have a geographical location. Looks very simple, and so it is, but there is one thing that makes it different from an achitectural drawing. Look at the moving numbers in the lower right corner: those are coordinates, and coordinates are the basis of mapping. They are somewhat different from the coordiantes you use in architectural drawings: there every drawing has its own coordinate system. Maps have in principle only one coordinate system: that of the globe You probably all know it: Every map in the whole world is in principle based on this system, and from this follows the following first principle of mapping: Every map can be combined with every other map. Of course, in reality things are never as easy as they look: maps are almost never designed in latitude-longitude coordinates, but in different projections, the earth isn't round etc. Yet in principle, every map can be transformed, so as to fit precisely over another map. That is the big difference with architectural drawings, and I am goint to explain to you how these drawings can also be brought into a mapping system.

The most important thing to know is that of projections The main thing to know is that a round surface like the earth can never be displayed on a flat surface like paper map. I'll illustrate the way a projection is made with the way Dutch maps are made, the Stereographic Projection If we show Europa in this projection, we see Europa from the viewpoint of the Netherlands. Distortion is small. However, if we look at the whole world, as seen with the Netherlands as center, we see a considerable distortion. Please note that every country has its own official projection: each county portrays itself as the center of the world. Even with small countries as Holland and Belgium this leads to small deviations16_nl_be.jpg")> Holland and Belgium this leads to small deviations. You can very clearly see this in this small application I guess the problems from the Rhine-Meuse maps come from this principle.

OK, what projection does Google use? Google is world wide, doesn't have a real center. As you see, Google maps are really different from national maps. Well, Google uses one of the oldest projections there is, the Mercator projection. Here is how it works. The problem is even more complex as the earth is not a sphjere and not even an exact ellipse: correct Boundary between Holland and Belgium in Southern Limburg, and incorrect These differences are small, but especially on topographical level they are signifivant, and that is the level you are working at.

Well, let's get practical: how do you get your architectural drawings into Google? Let me show you how I did it with the: TMK
1
2
3
4