slide visualization design for choropleths in multi scale statistical mapping

Multi-scale map Same theme created and mined/explored on screen screen maps Different visualization among predefined scale range... THEORITICAL FRAMEWORK AND METHODS 1 Rules for level

VISUALIZATION DESIGN FOR

CHOROPLETHS IN

MULTI-SCALE STATISTICAL MAPPING

Van Ngoc Truc Phuong, Le Minh Vinh

Faculty of Geodesy and Cartography,

HCMC University of Natural Resources and Environment

• Rationale

• Theoretical framework and methods

• Results and discussions

• Conclusions

Modern cartography takes new missions

which include building new principles that ensure thoughtful making and effective

use of map products from new and

innovative technologies

Multi scale maps

Multimedia

Multi-time

Maps on screen (screen maps) Viewed from distance of 80cm

Multi-scale map

Same theme

created and mined/explored on

screen (screen maps)

Different visualization among

predefined scale range

How to setscales’ threshold?

At each threshold, how do data classification and graphics change?

THEORITICAL FRAMEWORK

AND METHODS

THEORITICAL FRAMEWORK

AND METHODS

(1) Rules for level of scale range:

Based upon ability of human eyes to

THEORITICAL FRAMEWORK

AND METHODS

Zoom out Map

Map at smaller scale

readable smallest area unit <

Map at smaller scale

Classification change/ graphic transition:

move to higher administrative unit

readable smallest area unit

< 2.5

Cấp TK = Tỉnh/TP

Bản đồ cấp Tỉnh/ TP

N

N

Y

THEORITICAL FRAMEWORK

AND METHODS

Conventional map: best classification

methods ensure that all thresholds

have to make sense; all group

members concentrate around

representative values (if any) and have

appropriate GVF, GADF…

Multiscale: classification satisfy traditional rules and they also need to be consistent throughout the scale ranges

-> the classes for the less detailed mapping level should be translated to more detailed levels and add new more classes.

(2) Rules for data classification

THEORITICAL FRAMEWORK

AND METHODS

(3) Rules for color scheme

Conventional map: color ramp need to

present both hierarchical order and

have to corresponding to value of each

group (Use histogram)

Multiscale map: guarantee

consistency in color ramp among classes of different administrative unit levels

Case study: Layers of :

RESULTS AND DISCUSSIONS

Case study: Layers of :

RESULTS AND DISCUSSIONS

Case study: Layers of :

RESULTS AND DISCUSSIONS

RESULTS AND DISCUSSIONS

(1) SCALE RANGE

readable smallest area

unit

PROVINCE level: Bac Ninh 823,710249 1/18.000.000

District level: District 4 4,1593463588 1/1.300.000

Commune level: Ward 3,

classification is effected by various factors

Of which, value of classes have to satify

classification goals:

(a) to explore urbanization intensity

(b) to explore population distribution

RESULTS AND DISCUSSIONS

(2) DATA CLASSIFICATION

2a to explore urbanization intensity : criteria based

Type 2 10.000 x

Type 3 6.000 x Type 4 4.000 x

RESULTS AND DISCUSSIONS

(2) DATA CLASSIFICATION

Commune District Province

2b to explore population distribution:

Graph data distribution and pay attention to mean

values:

- Mean of Vietnam (265 person per/ sq km)

- Mean of HCMC at district level (3.589 per/sq km)

- Mean of peri-urban communes (1.075 per/sq km)

- Mean of inner communes (23.932 per/sq km)

- Mean of HCMC at commune level (28.293 per/sq km)

RESULTS AND DISCUSSIONS

(2) DATA CLASSIFICATION

Commune District Province

RESULTS AND DISCUSSION

(3) COLOR SCHEME

Explore data distribution for a corresponding color ramp

RGB Color RGB Color RGB Color

RGB Color RGB Color RGB Color

To have good digital map products, it is

necessary to stick to conventional map

principles and rules However, that’s not all/enough New principles and rules need

to be established These are new missions for modern cartographers

In this paper, we have set new rules for choropleths

in multi-scale statistical mapping The new rules set to be satisfied to the “multiscale” aspect, which are:

- Scale’s thresholds: defined by readable smallest area unit

- At a new scale range, different hierarchical level of administrative units will be displayed

- Classification and color scheme: be consistent throughout the scale ranges

Limitation of the paper:

although new rules are set, the maps in this

paper was made semi- automatically Follow-on research should build tools for automatic steps, such as tool for classification of multiscale data, use different Color Specification Systems for

better choice of color ramps

THANK YOU FOR YOUR ATTENTION