Sunday, March 05, 2006

The ionosphere influence in GPS precision

Along the path between the satellite and the receptor, GPS suffers different types of influence, either in speed propagation, polarization, signal power, direction, etc. This influence causes errors. One of the major factors that limits the positional accuracy provided by GPS, is the delay caused by the ionosphere to the passage of the signal. This is due to the elevated number of free electrons, excitated by sun light (night readings are more feasible), in this specific layer of the atmosphere.

For the GPS signal, the ionosphere is responsible for errors that can vary from 1 meter to more than 100 meters. For being a dispersive environment regarding the frequencies used in GPS, the ionosphere produces an effect to the signal, depending on the signal frequency and the refraction index. This effect is proportional to the TEC (number of electrons along the way between the satellite and the receptor).

If the value of TEC had a regular variation, the effects caused by the ionosphere would be of easy determination.

The problem is that TEC varies in time and space, by reason of the ionization solar flux, electromagnetical activity, spot solar cycle, station year, user location and direction of the satellite vector ray, making the process of ionospheric correction very difficult.

Nevertheless, the majority of the errors associated with the ionosphere can be eliminated through the observation of linear combination obtained in two different frequencies, remaining only the residual error. The problem is that most of the users work with one frequency receptors (much cheaper than double frequency receptors), taking no advantage of this fact. Also, the use of GPS in differential mode (DGPS) reduces the effects on the ionosphere, but this process is limited by the area where it can be used.

Therefore, it can be concluded that there is a need to develop models (e.g., Ionospheric Plate Model, Daily Cosine Model, Klobuchar Model) that eliminate the effect of the ionosphere in GPS readings, allowing this way, great positional accuracies.

António Araújo

The bibliography and full article is presented in this blog with the name "Modelling the Error Introduced by Ionosphere in GPS Signal in Receptors of One Frequency".

Geoid Ondulation

When we talk about the shape of the earth we can refere it under two different concepts. The first one, concerns to the description of the geometric physic surface and the second one to the shape of the equipotencial real gravitical field, being this one, important to the characterization of its properties.

One of this equipotencial surfaces is particularly significative, namely the one that is coincident to the free surface of the oceans (discounting the meteorological effects) denominated of Geoid.

Geoid is an equipotencial surface of the earth gravitical field, being close to the average level of the ocean and extended hypothetically under the oceans. It is also designated of equipotencial surface of level zero (surface where the gravitical force has a same intensity in all it’s extension).

With the arriving of GPS (Global Position System/Navigation System with Time
And Ranging), the knowledge of the geoid as taken a crucial role in the vertical positioning, more specifically, becomes to exist the need to transform the ellipsoidal altitude obtained in GPS into an ortometric altitude. In this transformation, the knowledge of the geoid, acquires a crucial importance.

Having the geoid an irregular shape, and not being, therefore, passible of become mathematically represented, it’s easier then, to obtain the position of each physical surface of the earth relating to a mathematical figure. That figure is in practice an ellipsoid which get’s so close to the shape of the geoid, that it’s difference rarely exceeds 100 meters, Nevertheless, each country has to adopt has reference the most adequate ellipsoid to it’s region, since there are different values for the several elements of the ellipsoid regarding it’s spatial location.

The point where the surface of the ellipsoid touches the earth, it’s given the name of datum, which characterized starting from a surface of reference (horizontal datum) and a level surface (vertical datum – geoid).

The geoid ondulation (N) is the distance between the geoidal surface and the ellipsoidal surface, counted along the plumb-line. This height is approximately equal to the difference between the ortometric (H) and ellipsoidal (h) altitude. The geoid ondulation differs on world from 107 to 85 meters, relating to ellipsoid of reference (WGS84). Therefore, can be written the following expression:


Since the process of geometric levelling is quite expensive and takes lots of time and GPS has proved to be very useful in spatial positioning, there was the need to also obtain H using this system. Nevertheless, as the altitudes obtained with GPS (h) refers to the surface of the ellipsoid and not to the geoid, it is necessary to know N to get to H.

Dora Neto

Bibliography and full version of this article is presented in this blog with the name of "Geoid Ondulation".

Saturday, March 04, 2006

GIS in Future

These days, it's important to prospect about the future since the investment made in some product or application, must be the most useful as possible. Therefore, what are in my opinion the big tasks to become accomplished in a GIS software in future?

I believe GIS is going to have a huge impact when finally provides a full 3-D environment allowing developers to usufruct of it's framework. There are being developed algorithms that helps optimizing spatial analysis tasks in a 3-D environment, but I believe that with the current computers speed, only in the next years will achieve a new GIS platform.

As comunicating with XML (Extensive Markup Language) as assuming a huge rule on Web Services using an internet protocol like for instance http, I believe that even more we'll replace desktop application's for web browsers. Nevertheless, desktop applications we'll allways be needed when running complex processes wich take to much time to process spatial data.

Another interesting task, is to provide a browser interaction with the features, replacing the image for a vector graphic. ESRI as announced that will support Scalable Vector Graphics (SVG) on top of the ArcGIS 9.2 software platform(new release). SVG is a language for describing graphics in XML.

Finally, a little comment about open source GIS. In the last 5 years, open source as been growing and assuming an important role on the GIS community. I believe it will still continue to grow and will provide in the next 2 years all basic GIS functionality. Nevertheless, don't be mistaken, the choice for an open source GIS product is a very difficult choice, and most of the cases leads us to an obsolete platform.

While on standard IT technology we still count on the millions of open source developers around the world, when it comes to specific GIS functionalities, it's hard to get things done. The choice for an open GIS software must be made very carefully understanding the advantages and disadvantages that the platform will provide specially when seeing in a long term.

GIS Today

As today we're getting used to see GIS applied essentially in displaying maps, we must realize that we'te not using the full potential of GIS capabilities. It is also wrong to assume that this is what gives more profit.

A geographic information system can provide full capabilities in such different areas (Remote Sensing, Urbanism, Medicin, etc) that to limit it in browsing maps sounds sometimes offensive.
Modelling reality phenomenons like traffic noise, predicting fire evolution or even the spills of petroliferous oil are some of the important topics that we can study using GIS.

Nevertheless, here in Portugal the investment in this area is very poor and when it happens it's wasted in an isolated information system with no future of becoming updated. There isn't today a conscience of building a GIS model to be used in a long term.

We have today several complex application's built in ArcView GIS 3.x (ESRI) that needs an investment to become updated to the new object models. There is a huge effort to be done here.

At 1990, since the appearence of object oriented programming (OOP), it was clear that OO (object-oriented) language had arrived to stay. But, even so, as usual, many application's where built knowing that they would become exceeded in a quite few years.

Four years ago, with the first appearence of ArcGIS 8.1 wich provided also a full framework based on OO, ESRI gave a big step forward in GIS software, assuming clearly as being the leader in this area. Intergraph and some other GIS companies loses here some space.

ESRI today, provides a complete GIS software in almost all areas, supporting even browser. The top products for web developer are ArcIMS and ArcGIS Server. The information about each one of this products will be available in this blog very soon. I'll even compare each product and show the differences between them and the advantages of each one.

To use GIS locally, ESRI provides ArcGIS 9.x as a full platform, allowing the user to have a full package of GIS sofwtare. Detailed information about this products are available at www.esri.com. Also on market, GeoMedia (Intergraph) also provides a great environment to realize some complex GIS tasks.

What is GIS

There are a lot of complex definitions about what GIS is. The simplest definition is that GIS is a software that allows to interact with spatial georeferenced data. This means that in theory we can represent almost everything we see each day in a GIS software.

In practice this is much more difficult as it might see. Indeed, modelling spatial information depends most of the cases in the available data and the scale that they can be used to represent some real model.

Generally speaking, only today, major companies and public sectors realized the importance of having georeferenced and structured spatial information as the need of exchange information as become more crucial to help providing good decisions. Nevertheless, there is still a big step to be made when it comes to create a unified information system between major map suppliers. This would allow the exchange of information in real time, without having to follow the constant bureaucracy that makes impracticable a fast take of action when an incident occurs. This could be done with a simple protocol between the several entities.

As Google showed to the common user, GIS can be very useful to provide spatial information that helps us knowing where we are relating to the world. As this information becomes more important to the people, several companies are trying to conquer the market of GIS (phone companies, etc.) and the need of having accurate spatial information as become then even more important.

The only way of representing geographic information, is to provide point coordinates that can relate the spatial information with the world.

To simplify I'll divide the coordinate systems in two major groups (geographic coordinates):
1. Ellipsoidal Coordinate System - Using an ellipsoidal surface, is defined a coordinate system with a latitude (defined relative to the equatorial plane of the ellipsoid), a longitude (defined relative to the meridian) and an ellipsoidal height (measured under a normal to the ellipsoid).
2. Cartographic Coordinate System – When using a projection to define a bidimensional coordinate system.

In order to get more precise coordinates, it is common to use an ellipsoid of reference, positioned in a earth station (local datum), whose natural coordinates are well known.