217 
MODELING OF CHANGE-ONLY INFORMATION AND UPDATING OF SPATIAL 
DATA 
YING shen 
School of Resource and Environment Science, Wuhan University. Wuhan .430079 .China - shy@whu.edu.cn 
KEY WORDS: Geographical Iinformation Science, Spatiotemporal Modelling, Geo-information Management, Database Updating, 
Change Detection 
ABSTRACT: 
We have constructed multi-scales foundational geographical database in china, and have released diverse spatial data and derived 
products. But how to keep the currency and continual updating of spatial data becomes the key problem. Now we almost update 
geographic information in batch volume mode. Data vender provides updated data to client users and the users update their own 
database according to the updated data. But this approach shields the concrete contents of geographic feature changes, as well as the 
difference of semantics, structures of geographic information between the vendors and clients. So batch updates method is time- 
consuming and troublesome with low efficiency. Incremental updates based on change-only information provide a novel approach 
for geographic data updating. Incremental updates only record the change-only information and characteristics in master database 
between different version datasets, and deliver the delta information to client users and integrate them into client database. The paper 
develops the technique workflow of extraction and distribution of change-only information based on different version databases. We 
take road features as an example to analyze the change types of realistic entities and discuss the object match method in the process 
of extraction of change-only information. Also we give the definition of change-only information and the idea about distribution and 
issue of change-only information to update user’s database. 
1. INTRODUCTION 
With the rapid development of economy and society, land 
covers and landscapes take place huge changes. Geographic 
information not only has spatial location characteristics, but 
also the temporal characteristics. On one hand geographic 
information records the entities of earth surface, on the other 
hand it reflects the currency and dynamics of geographic 
features, which make the updating and maintenance of spatial 
database to be arduous tasks. Because of the huge investments 
and complex specifications of process and updating of 
geographic data, we set up many professional companies and 
official sectors to manage and deal with spatial data, such as 
china state bureau of surveying and mapping (SBSM), NIMA, 
USGS, TeleAtlas. Up to date, china has taken shape of 
foundational geographic information and databases, has 
established multi-scale, diverse products of spatial data. But 
how to keep up with the currency of foundational spatial 
databases and continuous updating have become the hotspots 
and difficult problems (Chen Jun, et al. 2003). 
The future geographic information services will be “on demand 
and timely updating”, and they include change monitoring, 
change detection, database updating and products renewing. 
(Chen Jun, et al. 2007). In the environment of geographic 
information share, the data vendors with master database 
provide spatial reference data to other organizations and users, 
while the client users process and adjust geographic data to 
integrate with their own database according to their 
requirements. In order to keep the currency of spatial data, the 
client users must purchase the data in time from the vendors if 
they can provide. This has to meet with the problem of discover 
the change in master database and update the client database. 
The common method is batch updating, which has drawbacks of 
long period, high cost, and actually there are only small data 
have changed in the holistic area. Also the client users must 
face the risk of inconsistence and lost of their own geographic 
data with batch updating. Change-only updating gives a novel 
idea to update geographic database. It detects the difference 
between different version data in the master database and only 
records the changes information of geographic features and the 
corresponding characteristics, and delivers the change-only 
information to client users with delta files to fusion with client 
database. The change-only updating has the advantages of high 
efficiency, high speed and only takes place on limit geographic 
features with small volume data, which put change-only 
updating to the front hotspot. 
Now china has built foundational geographic databases with 
different versions. We develop the overall technique framework 
of detection of change-only information and change-only 
updating based on versions database to detect, extract, manage 
and publish the change-only information efficiently, which is 
illustrated in section 2. Geographic data represent the realistic 
geographic entities, so we will describe the changes in reality 
and digital environment in section 3. The paper takes the road 
feature as example, discusses the change taxonomy based on 
version database. In section 4 we develop the match approach to 
extract the change-only information. Also we give the ideas of 
modelling, transmission, updating integration of change-only 
information. 
2. FRAMEWORK OF CHANGE-ONLY INFORMATION 
AND UPDATING 
Here we define the difference between the new version database 
and old version database as change-only information. Generally 
we can produce the change-only information by subtracting the 
old database from new database and the mutual comparison. 
Detection and representation of change-only information is the 
key of change-only updating. To get change-only information