Read and access IIASA's models, tools, and data.

Research Projects

Read about individual research projects and activities at IIASA.

Research Partners

Read how IIASA works with research partners across the globe to deliver world-class research.

Models, Tools & Data

Read and access IIASA's models, tools, and data.

Contacts

ESM Program

Research Themes

Models/Tools/Data

Research Projects

Staff & Contact

YSSP Program

Geo-Wiki

The Geo-Wiki Project is a global network of volunteers who wish to help improve the quality of different thematic datasets globally.

Background

Land cover derived from remotely sensed products is an important input to a number of different global, regional and national scale applications including resource assessments and economic land use models. During the last decade three global land cover datasets have been created, i.e. the GLC-2000, MODIS and GlobCover, but comparison studies have shown that there are large spatial discrepancies between these three products. One of the reasons for these discrepancies is the lack of sufficient in-situ data for the development of these products.

In 2009 IIASA, the University of Applied Sciences Wiener Neustadt , and the University of Freiburg developed and set up a geospatial Wikipedia based on the Google Earth platform to address this issue; a crowdsourcing tool called Geo-Wiki has been developed. Geo-Wiki development has been financed by the European Commission via framework funding and an Austrian Funding Agency project called LandSpotting.

Geo-Wiki has two main aims:

To increase the amount of in-situ land cover data available for training, calibration and validation, and
To create a hybrid global land cover map that provides more accurate land cover information than any current individual product.

FAST FACTS

Geo-Wiki is part of a growing trend to reverse the usual top-down flow of data by allowing Internet users from any region of the world to be involved in a global land cover validation exercise.
IIASA has combined existing cropland data sets from sub-Saharan Africa to create a new map with higher accuracy than existing maps. The cropland map, which has an estimated 83% accuracy, is available for review.
Working with the TU Vienna, we are developing a series of games linked to Geo-Wiki as an incentive to volunteers to provide information.
IIASA is developing a robust methodology to improve the quality of the validations provided by volunteers.

How Geo-Wiki works

Volunteers are asked to indicate the land cover type of a given area based on what they can see on Google Earth. Their input is recorded in a database, along with uploaded photos, to be used in the future for the creation of a new and improved hybrid products.

Once in the system, Google Earth is displayed with the in-built functionality to rotate the Earth and to zoom into any land surface to begin land cover assessment. The user can also plot any of the global land cover maps (GLC-2000, MODIS, GlobCover) on top of Google Earth, or the user can display the disagreement maps between any pair of land cover products . These disagreement maps highlight the critical areas where further validation is needed.

It is these large areas of disagreement that Geo-Wiki is targeting in its crowdsourcing efforts.

Click here for more information on how to assess land cover.

Geo-Wiki Branches

AGRICULTURE.Geo-Wiki.org: users are asked to determine the percentage of agricultural area that can be seen in a MODIS-sized pixel using Google Earth. In addition to the global land cover products GLC-2000, MODIS v.5 and GlobCover 2005, different global and regional agricultural datasets can be viewed
BIOMASS.Geo-Wiki.org: A collection of global, regional and in-situ biomass datasets designed for comparative analysis and validation
HUMANIMPACT.Geo-Wiki.org: Users are asked to determine the degree of human impact and agricultural intensity that is visible in a defined area using Google Earth. Additionally users are asked to provide the land cover type and the level of abandonment.
REGIONAL.Geo-Wiki.org: A set of landcover products for Russia including carbon budget parameters.
URBAN.Geo-Wiki.org: Users are asked to determine the percentage of urban area that can be seen in a MODIS-sized pixel using Google Earth. Urban areas can be viewed for the GLC-2000, MODIS v.5 and GlobCover 2005. Disagreement maps between each pair of land cover products can also be displayed.
COMPETITION.Geo-Wiki.org: Owing to the large amount of disagreement among global land cover products, the creation of a global hybrid land cover product is seen as a major step towards improvement. In that regard, the goal of this competition is to focus on the areas of greatest disagreement among the existing products.