METHODOLOGICAL QUESTIONS THE LAND COVER CHANGE OF THE URBAN AREAS IN SZEGED (HUNGARY) CASE STUDY

L. MucsiUniversity of Szeged Dept. of Physical Geography and Geoinformatics

AIMS of the STUDY

Spectral information

Textural information

SMA NDVI

Segmentation

Statistical parameters

Land cover map

Land use map

VISmodel

Land cover change detection (1987‐2011)

UHI (1987‐2011)

Condition of green areas

Demography

DATA, METHODS AND APPLICATIONS

• Long term land use and land cover change– Landsat TM images– Pixel based and subpixel classification– Application: Urban heat island mapping

• Annual change– Vegetation mapping

• Large scale mapping– Rapideye image– segmentation

Functional districts of Szeged

Land use and land cover change

Landsat TM, ETM+ 453 (RGB)

1986. 06.19. 1991.06.17. 2002.06.23. 2006.09.30.

RATIO MAPS FOR LAND COVER CLASSES USING NSMA METHOD

• 3 endmembers (impervious surface, vegetation, soil) spectral unmixing(10th July 2011, Landsat TM image)

Accuracy assessment

• 250 samples 90 m × 90 mRMSE=17 és SE=‐4,1

38,3% 41,4%

N

)VV̂(RMSE

N

1i

2ii

N

)VV̂(SE

N

1iii

Connection between land cover and land use 

1 – Forest, 2 – Rare vegetation 3‐ Soil with vegatation cover, 4 – Bare soil surface, 5 – Grassland, 6 – low density build area, 

7 – medium density build area, 8 – high density build area, 

9 – commercial zone

Impervious surface‐Vegetation‐Soil compositeModified VIS diagram

PSAK XI. Konferencia, Szeged 2012.11.07-09.

11

Annual change of NDVI values

Vegetation

Soils

Impervious surfaces

ND

VI v

alue

s

No. of days in a year

Standard deviation map made of 8 NDVI images from the time series of 1986

Fraction rate map of artificial surface taken from the spectral partition of a Landsat TM satellite 

image (24th July 1987)

Relation between the standard deviation of NDVI values and the rate of impervious surface

Fraction map of artificial surface calculated from the deviation map

8 days air temperature measurement

‐2‐10123456789101112131415161718192021

1 15 29 43 57 71 85 99 113

127

141

155

169

183

197

211

225

239

253

267

281

295

309

323

337

351

365

379

393

407

421

435

449

463

477

491

505

519

533

547

561

575

589

603

617

631

645

659

673

687

701

715

729

743

Hőm

érsé

klet

 (°C)

5037 ‐Baktói kiskertek

5047 Radnóti Miklós kísérleti Gimnázium

5049‐Tömörkényi  István Gimnázium

5066 Alsóváros

THE SPATIAL DISTRIBUTION OF AVERAGE MAXIMUM UHI INTENSITY VALUES IN THE NON‐HEATING SEASON 

IN 1986 and 2011The maximum UHI intensity (∆T) can described as a function of build

up ratio (B) with the best fit regression line in the non heating period of the year (from April to October) with the following equation:

∆T = 0,018B + 0,716

RapidEye satellite image in the range of visible light recorded March 24, 2011

Figure 1.

PSAK XI. Konferencia, Szeged 2012.11.07-09.

20

Result of object based classification

• Segment (object) was classified according to the majorityof classified pixels inside the segment

Average Shape Index and Pattern Size of  Funtional Districts

23

Downtown, inner residential and housing estate districts in the plane spanned by mean patch size and mean shape index

Conclusions• After radiometric and atmospheric correction, the time

series analysis of LTM images can give usable statistical parameter for assessment of impervious surfaces in each pixel in urban region

• Standard deviation of annual NDVI values is in strong correlation with the fraction values of impervious surfaces in NSMA

• Intensity of average annual maximum UHI values can be calculated from the fraction of impervious surfaces, so from the SD of annual NDVI values

• Combined geometrical and spectral parameters can be used for characterize segments and for supervised classification which resulted pixel based thematic map

• The thematic map is the input for segment object based classification, which resulted urban land cover map of high accuracy

Thank you for your attentions!