Directeur de Thèse

Christiane Weber

Description

Niveau: Supérieur, Doctorat, Bac+8
Directeur de Thèse : Rapporteur Interne : Rapporteur Externe : Rapporteur Externe : Examinateur : Mme Christiane WEBER, Directeur de Recherche CNRS Université Louis Pasteur I M Dominique SCHWARTZ, Professeur Université Louis Pasteur I Mme Laurence HUBERT-MOY, Professeur Université de Rennes II M Jürgen BREUSTE, Professeur Universität Salzburg, Autriche M Michael BRUSE, Professeur Universität Mainz, Allemagne Thèse présentée pour obtenir le grade de Docteur de l'Université Louis Pasteur Strasbourg I Discipline : Géographie par Annett Wania Urban vegetation – detection and function evaluation for air quality assessment Soutenue publiquement le 12 Novembre 2007 Membres du jury

• french urban


• deeply grateful


• dong binh


• annett wania


• universität mainz


• function evaluation


• rapporteur externe


• urban vegetation


• directeur de la recherche

Sujets

• Schwartz
• Louis Pasteur University
• Lennon
• Pasteur
• Strasbourg
• Hirsch

Informations

• Publié par: profil-zyak-2012
• Publié le: 01 novembre 2007
• Nombre de lectures: 47
• Langue: English
• Poids de l'ouvrage: 28 Mo

Extrait

Thèse présentée pour obtenir le grade de

Docteur de l’Université Louis Pasteur

Strasbourg I



Discipline : Géographie
par Annett Wania
Urban vegetation – detection and
function evaluation for air quality
assessment


Soutenue publiquement le 12 Novembre 2007
Membres du jury

Directeur de Thèse : Mme Christiane WEBER,
Directeur de Recherche CNRS
Université Louis Pasteur I
Rapporteur Interne : M Dominique SCHWARTZ, ProfesseurI
Mme Laurence HUBERT-MOY, rRapporteur Externe :
Université de Rennes II
M Jürgen BREUSTE, Professeur
Universität Salzburg, Autriche
M Michael BRUSE, rExaminateur :
Universität Mainz, Allemagne








It is particularly ironic that the battle to save the world's remaining
healthy ecosystems will be won or lost not in tropical forests or coral
reefs that are threatened but on the streets of the most unnatural
landscapes on the planet.
(Worldwatch Institute 2007)



Acknowledgments
This thesis was developed and authored at the ‘Laboratoire Image et Ville’, Department of
Geography and Planning at the Louis Pasteur University in Strasbourg, France.
First of all thank you Christiane Weber, head of the ‘Laboratoire Image et Ville’, for guiding
me through this thesis and sharing your insights and views on the French urban ecology. I am
deeply grateful for the opportunity you gave me to live and work in France.
Thank you Prof. Laurence Hubert-Moy, Prof. Jürgen Breuste, Prof. Dominique Schwartz and
Prof. Michael Bruse for investing your time and effort in the evaluation of this thesis.
Many thanks to you, Michael Bruse, for your patient help and discussions concerning the
ENVI-met model and for your insights on atmospheric modelling.
Thank you Nadège Blond for your discussions and critical comments on the air quality
simulations.
Thanks, Alexandre Blansché, for processing my data using MACLAW and the opportunity
you gave me to view the process of image analysis through the eyes of a programmer. In this
context thanks also to Pierre Gançarski for his insights.
For your input on remote sensing, thank you Jacky Hirsch, Aziz Serradj, Dong Binh Tran,
Stéphane Lhomme, and Marc Lennon.
Furthermore, I thank ‘l’Association pour la Surveillance et l’Etude de la Pollution
Atmosphérique en Alsace’ (ASPA) for providing me the STREET database and
meteorological data as well as the ‘Service des Espaces Verts, des Jardins Familiaux et des
Forêts’ of the municipality of Strasbourg for providing me data from the tree database.
This thesis would not have been the same without my companions from room 417: Thank you
all for sharing your time, tea, and biscuits with me. Merci beaucoup au LIV!
For your support in verbalizing this thesis in proper English, I thank you Amy and Patrick.
I am particularly thankful to Florian for sharing with me the beauties of life…and the end of
this thesis. Thanks to my family for supporting me in my projects even from far away.
And finally, thank you FIP!


5
TABLE OF CONTENT
LIST OF FIGURES 9
LIST OF TABLES 13
INTRODUCTION 15
PART I: THEORETICAL FRAMEWORK 19
1 SUSTAINABLE DEVELOPMENT AND URBAN AREAS 19
1.1 Sustainable development 19
1.2 Urban areas –a chance for sustainability? 21
1.2.1 Urbanisation and city growth 21
1.2.2 A chance for sustainability? 22
1.2.3 Political action for a sustainable urban environment 23
1.3 Urban green and urban sustainability 25
2 URBAN VEGETATION – AN ELEMENT OF THE URBAN ECOSYSTEM 27
2.1 The System and Ecosystem Approach 27
2.1.1 The system approach 27
2.1.2 The ecosystem approach 30
2.2 The urban ecosystem 33
2.2.1 Approaches to urban ecology 34
2.2.1.1 Urban ecology in the natural sciences 34
2.2.1.2 Urban ecology in the social sciences 34
2.2.1.3 System analysis 35
2.2.1.4 Conclusions and adoption of one approach 36
2.2.2 Characteristics of urban ecosystems 38
2.2.2.1 Structure and metabolism 38
2.2.2.1.1 Boundaries 38
2.2.2.1.2 Elements 38
2.2.2.1.3 Metabolism 39
2.2.2.2 Physical environment 43
2.2.2.2.1 Soils and water 44
2.2.2.2.2 Atmosphere 45
2.2.3 Conclusions 51
2.3 The ecosystem element vegetation 51
2.3.1 Definitions and terms 52
2.3.1.1 Botanic terms 52
2.3.1.2 Terms used in planning 54
2.3.1.3 Concluding remarks 57
2.3.2 The element and its interactions 57
2.3.2.1 Interactions induced by the plant’s metabolism 58
2.3.2.2 Influence of humans on plants 59
2.3.3 Conclusions 62
2.4 Vegetation and urban environmental quality 62
2.4.1 Plants as indicators of urban environmental quality 62
2.4.2 Ecological functions 66
2.4.2.1 Climate 66
2.4.2.2 Air pollution 67 6
2.4.2.3 Water 69
2.4.2.4 Biodiversity 69
2.4.3 Social, psychological, physical, and aesthetical functions 70
2.4.3.1 Psychological, and physical functions 70
2.4.3.2 Social functions 70
2.4.3.3 Aesthetic functions 71
2.4.4 Economic functions 71
2.5 The need to observe and promote vegetation in cities 72
PART II: DETECTION OF URBAN VEGETATION 75
3 VEGETATION DETECTION USING REMOTELY SENSED DATA 77
3.1 Some basic principles of remote sensing 77
3.2 Spectral characteristics of vegetation 78
3.2.1 Influencing parameters 80
3.2.1.1 Leaf structure 80
3.2.1.2 Leaf pigments and water content 81
3.2.1.3 Bi-directional effects 82
3.3 Urban vegetation analysis 84
3.3.1 The use of remote sensing for urban vegetation detection 84
3.3.2 Studies on urban vegetation 85
3.4 Hyperspectral data for urban vegetation analysis 86
3.4.1 Characteristics of hyperspectral data 86
3.4.2 Methods in hyperspectral data analysis 88
3.4.2.1 Sub-pixel mapping 88
3.4.2.2 Whole pixel mapping 89
3.4.2.3 Dimensionality reduction 91
3.4.2.3.1 The problem of dimensionality 91
3.4.2.3.2 Feature selection 91
3.4.2.3.3 Feature extraction 93
3.4.2.4 Summary on performed dimensionality reduction methods 97
3.5 Urban vegetation analysis with CASI data 98
3.5.1 Data 99
3.5.2 Spectral properties of urban vegetation cover 102
3.5.2.1 Specific radiative features of urban areas 102
3.5.2.2 Vegetation specific features 103
3.5.3 Vegetation indices and variations in red-edge inflection point 108
3.5.3.1 Vegetation indices 108
3.5.3.1.1 Description of the used indices 108
3.5.3.1.2 Comparison of results 111
3.5.3.2 Variations in red-edge inflection point 112
3.5.3.2.1 Calculation of the red-edge 112
3.5.3.2.2 Results 113
3.5.3.3 Summary 114
3.5.4 Supervised classification to detect tree species 115
3.5.4.1 Definition of training sets 116
3.5.4.2 Feature reduction 119
3.5.4.3 Supervised classification 120
3.5.4.4 Performance of hyperspectral data compared with multispectral data 123 7
3.5.4.5 Conclusions 124
3.6 Test of a new algorithm for feature selection in hyperspectral images 125
3.6.1.1 Data and methods 125
3.6.1.2 Supervised classification on transformed features 126
3.6.1.3 Unsupervised classification and feature selection with MACLAW 128
3.6.1.4 Evaluation of the performance of MACLAW band selection 130
3.6.2 Conclusions 131
3.7 Discussion 132
PART III: EVALUATION OF THE FUNCTION FOR AIR QUALITY 135
4 MICROSCALE AIR QUALITY SIMULATIONS 135
4.1 Definition of a spatial scale 136
4.2 The ENVI-met model 137
4.2.1 Model description 137
4.2.1.1 General model design 137
4.2.1.2 Model configuration and model area definition 139
4.2.1.3 Some basic modelling principles 140
4.2.2 Model validation 143
4.3 Simulation of the influence of vegetation on particle dispersion in standard scenarios 144
4.3.1 Model configuration 145
4.3.1.1 Basic parameters 145
4.3.1.2 Vegetation scenarios 148
4.3.2 Flow and particle dispersion in the situation without vegetation 151
4.3.2.1 Wind flow 151
4.3.2.2 Particle dispersion 156
4.3.2.3 Conclusions 161
4.3.3 Influence of vegetation on particle dispersion 162
4.3.3.1 Influence on particle concentration 162
4.3.3.2 wind speed 166
4.3.3.3 Influence of wind speed on vegetation-induced concentration changes 171
4.3.3.4 Particle removal by plants 175
4.3.3.5 Conclusions and planning recommendations 179
4.4 Simulation of the influence of vegetation on air quality in a real case scenario 181
4.4.1 Model configuration 181
4.4.2 Results 182
4.4.3 Conclusions 185
4.5 Discussion 186
5 GENERAL CONCLUSIONS 191
5.1 Perspectives 193
REFERENCES 195
GLOSSARY OF TERMS 217
ABBREVIATIONS 221
EXTENDED SUMMARY IN FRENCH / RÉSUMÉ EN FRANÇAIS 223
ANNEX 235
8
9
LIST OF FIGURES

Figure 1.1: The three key dimensions of sustainable development and the main subjects to be addressed in
each dimension (Allemand 2006, IPCC 2007). 20
Figure 1.2: Urban population as percentage of total population for 2005 and predicted for 2030 (UN 2006). 21
Figure 2.1: Simplif