THE MULTI-RESOLUTION CHARACTERISTICS OF SPATIAL DATA IN VIETNAM 
LAND ADMINISTRATION 
T. N. Trung* 
* IMECOSUM, Ministry of Natural Resources & Environment, 71 Nguyen Chi Thanh, Hanoi, Vietnam — 
TrungGeomatics@Pmail.vin.vn,Tran_Nhu_Trung@Y ahoo.com 
Commission IV, WG IV/3 
KEY WORDS: Analysis, Acquisition, Generalization, Multiresolution, Management, Land Use 
ABSTRACT: 
Corresponding to the characteristics of multi-discipline and multi-level management, land administration always requires and 
acquires spatial data at very different spatial and thematic resolution. This particularity leads to many problems in spatial data 
acquisition and data management such as: inconsistency and difficulty in exchanging data between data sources and application 
disciplines. There are many researches in the direction of multi-resolutions to solve above questions. However, the adequate solution 
only comes out when the multi-resolution characteristics are quantified for a certain application. This paper has the objectives to 
quantify the multi-resolution characteristics of spatial data in Vietnam land administration. This quantification bases on the analysis 
of function, role of each discipline and administrative management level for land administration system. To verify the use of the 
quantification results, a case study of generalization, acquisition of land use data between disciplines and management levels is 
carried out at a local level. 
1. INTRODUCTION 
1.1 General introductions 
Spatial data (SD) is the core data for land administration (LA) 
activities. Vietnam LA is implemented at four levels, namely 
national, provincial, district and communal level (so called 
multi-level characteristics). There are many disciplines 
involved in LA activities such as land use planning, land 
registration, land valuation, etc (so called multi-discipline 
characteristics). 
Corresponding to the multi-discipline and multi-level 
characteristics, LA requires and acquires spatial data at 
different spatial and thematic resolution. This particularity leads 
to many problems in SD acquisition and SD management such 
as: inconsistence, in-accuracy, difficulty in exchanging data 
between sources and disciplines. 
Vietnam LA witnessed many cases of inconsistent SD supplied 
by different management levels and different disciplines. A 
typical example is the case of land use data supplied by land 
use statistics and land use mapping in 1995. Recently, in the 
workshop of preparation for land use inventory for the year 
2005, many opinions claim that land use information 
generalized from communal to national levels might have 
information accuracy between 50-70%, (RSC-MONRE, 2003). 
Solving the above practical problems are the objectives of many 
GIS researches working on multi-scale, multi-resolution and 
multi-representation direction. Regarding data model, database 
framework, one can find out several researches as follow. 
(Skogan, 2001) presented a framework for multi-resolution 
object database, which can be used inside a multi-resolution 
database. (Zhou and Jones, 2001, 2003) worked out a multi- 
representation data model and database. Specially, project 
MurMur (Parent, 2000) shows an integrated result in multi- 
resolution research with a data model which allows to work on 
the current commercial GIS, DBMS. MurMur develops a more 
flexible representation schemes, which allow end-user to may 
manage easily information representation. — (Kavouras and 
Kokla, 2002) proposed an integration of both vertical axe 
(between scale — resolution) and horizontal axe (between 
applications) for a multiple applications and multi-scale data 
required. 
However, the adequate solution comes out only for a certain 
application when the multi-resolution characteristics are 
quantified. Different context requires SD at different 
resolutions. This observation is even more detailed in many 
works, which are synthesised by (Molenaar, 1998). He 
considers the questions of multi-scale (as transferred to multi- 
resolution latterly) originated from the demand of research and 
manage the natural phenomena of the real world at different 
points of view. Analysis of different applications such as 
watershed management, land management, cartography, he has 
proposed four driven(s) for SD generalization, namely (1) class 
driven, (2) functional, (3) structure and (4) geometrical 
generalization. In short, the complete solution depends very 
much on the purpose of the application. 
For LA, there is a limited research touches inside of SD in the 
view of multi-resolution, except (Williamson and Fenecy, 2001; 
Rajabifard et al, 2000) when considering LA as a spatial data 
infrastructure, where multi-resolution or spatial reasoning 
hierarchy is obviously required. 
Particularly to Vietnam LA, the author of this paper has 
reported on the demand of multi-resolution of SD for multi- 
level land management. Consequently, there are several 
proposed solutions to exchange land use information between 
land registration and land use planning. Suitable data model to 
handle the uncertainty of land use classes of cach land use units 
are developed for Vietnam LA, see (Trung, 2001, 2002). 
238 
Internat 
Howeve 
complet 
four a 
resoluti 
each me 
This ] 
characte 
these r 
resoluti 
1.2 Me 
To qua 
analyse: 
and (2) 
level de 
guidelir 
will be : 
In the 
resoluti 
- Spa 
min 
- The 
- This 
reso 
To exp 
exchang 
and lan 
will be 
The par 
discipli 
are the 
requirec 
resoluti 
and dis 
section 
2. TH 
DI 
AL 
To bett: 
two vie 
docume 
discipli 
where t 
system. 
2.1 Th 
LA is d 
dissemi 
land w 
ECE, | 
country 
Willian: 
Four m. 
are illus 
levels c