Consultancy Study to Analyse Broad 
Land Use Pattern of Hong Kong

Executive Summary

Department of Geography
The Chinese University of Hong Kong
April 2001


1.  Introduction

2.  Review of Current Procedures for Preparing and Updating Land Use Data

3.  Land Use Classification System

4.  Technical Solution for Production of the Land Utilization Plan and Land Usage Table through Application of Remote Sensing Technique

5.  Future Updating of Land Utilization Plan and Land Usage Table

6.  Feasibility of Sale of Land Use Data

7.  Concluding Remarks

Table 1 Recommended Land Use Classification System

Table 2 Land Usage Table, 2000

Figure 1 Land Utilization Plan, 2000 (with reduced scale)


1.    Introduction

1.1

The objective of this consultancy study is to analyse the broad land use pattern of Hong Kong in order to facilitate a production of a small-scale (i.e. 1:75,000) Land Utilization Plan and compilation of the Land Usage Table of Hong Kong. With the availability of the latest aerial photographs and satellite images, the technologies of Remote Sensing and Geographical Information System (GIS) are applied in this study.

1.2

The Land Utilization Plan and the Land Usage Table of Hong Kong were first prepared by the then Town Planning Office of the Building and Lands Department in 1988 providing planning information to the public on the existing land usage of the Territory. At present, the Planning Department is responsible for the annual updating of the Land Usage Table, which is now attached at the Hong Kong Annual Report as an appendix. Due to physical and staff constraints, it was not possible to collect all updated data from field survey, in particular those for rural areas. Hence, the updating works undertaken were mainly based on data collected from various sources and compiled by different methods.

1.3

For the purpose of deriving a more systematic and efficient way to produce an updated Land Utilization Plan and a broad Land Usage Table of Hong Kong as well as facilitating the updating works in future, the study was commissioned to the Department of Geography, the Chinese University of Hong Kong .

    [top]


2.  Review of Current Procedures for Preparing and Updating Land Use Data

Current procedures used in generation of land use data within the Planning Department were reviewed.  The following features are identified :

(a) the Land Use Survey Team of the Planning Department has mainly been responsible for preparing large-scale (1:1,000) land use plans in the urban and new town areas;
(b) due to resources constraint, part of the land use data in rural areas cannot be frequently updated except for some specific land uses (e.g. port back-up or large development areas);
(c) land use classification schemes adopted vary among different planning studies; and
(d) most classification schemes do not adopt a hierarchical structure and thus render comparison among the classification schemes difficult.

     

 [top]


3.    Land Use Classification System

3.1 With reference to the current government practice in the Mainland and the USA, a new land use classification system with the following characteristics is recommended to facilitate integration of land use data derived from different sources and comparison among land use data:
(a) it is hierarchical with multi-level in structure to facilitate comparison and compatibility among land use data collected in different times and from different studies;
(b) it is designed based on the existing land use plans prepared by the Planning Department; 
(c) it takes into account the need to collect land use data from satellite images and aerial photographs; and
(d) it contains two levels of land use categories (i.e. Level-I and Level-II) to handle broad land use classification. Further dis-aggregation can be performed to provide more detailed classifications.

     

3.2    The recommended land use classification system is shown in Table 1.

Table 1  Recommended Land Use Classification System     [Simple Table Format]

Level - I Level - II
Code

Categories

Code

Categories

1

Residential Land

1.1

Private Residential

 

 

1.2

Public Residential

 

 

1.3

Rural Settlements

2

Commercial Land

2.1

Commercial/Business and Offices

3

Industrial Land

3.1

Industrial

 

 

3.2

Industrial Estates

 

 

3.3

Warehouse and Storage

4

Institutional Land

4.1

Government, Institution and Community Facilities

5

Transportation

5.1

Roads

 

 

5.2

Railways

 

 

5.3

Airport

6

Open Space

6.1

Open Space

7

Vacant Land

7.1

Vacant Development Land/Construction in Progress

8

Other Urban or Built-up Land

8.1

Cemeteries and Crematoriums

 

 

8.2

Public Utilities

    8.3

Others

9

Agricultural Land

9.1

Agricultural

 

 

9.2

Fish Ponds/Gei Wais

10

Woodland/Shrubland/Grassland

10.1

Woodland

    10.2

Shrubland

    10.3

Grassland

11

Wetland

11.1

Mangrove and Swamp

12

Barren Land

12.1

Badland

 

 

12.2

Quarries

    12.3

Rocky Shore

13

Water

13.1

Reservoirs

 

 

13.2

Streams and Nullahs

[top]


4.       Technical Solution for Production of the Land Utilization Plan and Land Usage Table through Application of Remote Sensing Technique

4.1 Taken into account the cost and the spatial resolution required for the Study, updated satellite images from LANDSAT-7 ETM+ and SPOT-2 HRV were adopted. In addition, ancillary data including digital maps, orthophotographs, digital elevation models and the existing land use plans were collected. The information were integrated and input to a computer image processing system for digital image analysis.
4.2 In the process of preparing the Land Utilization Plan and Land Usage Table, the entire territory of Hong Kong has been divided into three sub-areas with different methodologies adopted.  Details are as follows:
(a) urban areas (i.e. Hong Kong Island, Kowloon and Tsuen Wan) -- the existing and updated land use data were used with suitable cartographic generalization;
(b) new town areas -- as updated land use data for some areas were not available, change detection techniques were applied to highlight areas of change and facilitate updating of  the land use data; and
(c) rural areas -- satellite images are used for digital image classification.  Land use classes are generated based on the spectral reflectance of the associated land use and land cover classes.
4.3 Data derived from the three sub-areas were integrated for further editing and fine-tuning to produce the final Land Utilization Plan.  Accuracy assessment was undertaken to determine overall accuracy of the entire plan as well as accuracy of individual land use categories.
4.4 The Land Utilization Plan is produced at a scale of 1:75,000.  The area of the land use categories shown in Land Usage Table is computed based on the Land Utilization Plan.  The Land Utilization Plan (with reduced scale) and Land Usage Table are shown in Figure 1 and Table 2 respectively.

   Figure 1  Land Utilization Plan, 2000 (with reduced scale)

Land Utilization Plan, 2000 (with reduced scale)

 

Table 2  Land Usage Table, 2000

Land Use Categories

Approximate
Area (km2)
%

1       Residential Land    

   

1.1    Private Residential

24 2.2

1.2    Public Residential

16 1.5

1.3    Rural Settlements

 

27 2.4

2       Commercial Land 

   

2.1    Commercial/Business and Offices

 

3 0.3

3       Industrial Land

 

 

3.1    Industrial

5 0.5

3.2    Industrial Estates

3 0.3

3.3    Warehouse and Storage

 

11

1.0

4       Institutional Land

 

 

4.1    Government, Institution and Community Facilities

 

20

1.8

5       Transportation

 

 

5.1    Roads

35 3.2

5.2    Railways

1 0.1

5.3    Airport

 

13 1.2

6       Open Space

 

 

6.1    Open Space

 

20

1.8

7       Vacant Land

 

 

7.1    Vacant Development Land/Construction in Progress

 

27 2.5

8       Other Urban or Built-up Land

 

 

8.1    Cemeteries and Crematoriums

5

0.5

8.2    Public Utilities

6 0.5

8.3    Others

 

14 1.3

9       Agricultural Land

 

 

9.1    Agricultural

57 5.2

9.2    Fish Ponds/Gei Wais

 

13 1.2

10     Woodland/Shrubland/Grassland

 

 

10.1  Woodland

190 17.3

10.2  Shrubland

241 21.9

10.3  Grassland

 

310 28.2

11      Wetland

 

 

11.1   Mangrove and Swamp

 

6 0.5

12     Barren Land

 

 

12.1  Badland

16

1.5

12.2  Quarries

3 0.3

12.3  Rocky Shore

 

5 0.5

13     Water

 

 

13.1  Reservoirs

24 2.2

13.2  Streams and Nullahs

4 0.4

Total

1099 100.0

Remarks :
(1)  The area of mudflats is excluded.
(2)  The above figures are not added up to the totals as the figures have been rounded up.

[top]


5.    Future Updating of Land Utilization Plan and Land Usage Table

5.1 Future updating of the Land Utilization Plan can make use of the remote sensing change detection techniques. These techniques can easily highlight areas of change and facilitate the updating procedure.
5.2 It is, however, important to note the possible error propagation problems on continuing updating the land use database.  A comprehensive review on the Land Utilization Plan and Land Usage Table is recommended to be undertaken every five years though such frequency is subject to review if necessary.  In addition, in order to cater for the changing land use pattern, a regular review on the land use classification system is recommended.

     [top]


6.    Feasibility of Sale of Land Use Data

Based on the past sale figures, there is a demand for the Land Utilization Plan in both public and private sectors. A broad assessment of relevant factors including potential buyers, pricing and cost, copyright, internal resources implication etc., indicated that it is feasible to sell the Land Utilization Plan in both digital and paper format. However, further investigation would be required to resolve the technical and logistics issues.

[top]


7.    Concluding Remarks

7.1 This Study has accomplished its objectives in updating the Land Utilization Plan for the entire territory of Hong Kong at a scale of 1:75,000, from which the Land Usage Table is produced. With the application of remote sensing and GIS techniques, the broad land use data can be updated in a more comprehensive and scientific way.
7.2 With the advancement of remote sensing technology and the availability of high spatial resolution images (e.g. IKONOS satellite image with 1-metre spatial resolution) and new techniques in image processing analysis, more detailed information could be captured from the satellite images to facilitate the updating of land use data and better integration with the existing GIS databases. New applications and better technical solutions are expected to be available in the future. Review of the methodologies of land use data collection would therefore be necessary in due course.
7.3 Large amount of land use data can be derived from the image processing analysis. However, it should be noted that the task is labour intensive and a large portion of the processing still requires human decision. Staff experience in land use classification, mapping as well as conventional visual interpretation is of utmost importance in the process.

[top]