WP6 - Calibration/validation of the MOLAND model

Recently, spatial metrics have been introduced in the field of urban land-use mapping and modelling to characterize the spatial dynamics of such systems. In this work package it is studied whether spatial metrics can be applied directly on remote sensing images to calibrate and validate dynamic land-use models of urban systems.


More than ever before, planners and policy makers need tools to anticipate and assess the impact of their decisions on the spatial systems that they are to manage. A growing number of high resolution dynamic land-use models is currently being developed for this purpose.

Typically the required time series of land-use maps based on identical and consistent mapping methodologies, legends and scales are missing. As a result, the differences observed in a series of land-use maps are often not real land-use changes, but mismatches that should not be compared with model predictions of land use changes.

On the other hand, earth observation satellites provide images of the earth's surface since the early seventies. Their temporal availability is relatively high compared to land-use maps and a series of remote sensing images of one sensor is consistent in time and space. However, the spatio-temporal availability and consistency heavily depend on atmospheric conditions. The main problem is that conventional remote sensing based classifications result in land-cover maps, based on physical properties of the surface, rather than land-use maps representing functional classes.


The metrics developed in WP4 will be tested relative to their effectiveness for calibration and validation of the MOLAND applications of Dublin and Istanbul. To that effect their performance will be tested against classic historic calibrations for both cities.


At this moment land-use maps have been derived from remote sensing using metrics calculated from subpixel classification methods developed in WP2. First results have been obtained using metrics at the city block level and the kernel level which uses the Optimized Spatial Reclassification Kernel (OSPARK).
More info ...





Last modification date = 27-04-2009