Evolution and future projections of the urban heat island at the coastal urban area of Athens D. Founda 1, C. Giannakopoulos 1, M. Hatzaki 1, M. McCarthy.

Evolution and future projections of the urban heat island at the coastalurban area of Athens

D. Founda1, C. Giannakopoulos1, M. Hatzaki1, M. McCarthy2,C. Goodess3, D. Hemming2, and F. Pierros1

1Institute of Environmental Research and Sustainable Development, National Observatory of Athens, Greece2Met Office Hadley Centre, UK3Climatic Research Unit, School of Environmental Sciences, University of East Anglia, UK

• Athens is a large urban area of eastern Mediterranean that experiences both global warming and urbanization effects

• According to the historical records of the National Observatory of Athens (NOA), the annual air temperature in Athens reveals a statistically significant positive trend from the beginning of the past century till now

• This is due to the increase of the summer rather than the winter temperature

• The increase of the maximum temperature in summer during the last few decades (since the mid 1970’s) is significantly large and amounts roughly to 1°C/decade

• Summer 2007 was an exceptionally hot summer in Athens, with three severe heat waves and an all time record value in the maximum temperature (44.8°C at NOA)( Founda & Giannakopoulos 2009)

• However, summer 2012 (June – August) was the warmest summer ever recorded at NOA as regards the mean maximum and minimum air temperature

Air temperature trends in Athens

y = 0.02x - 4.61

24

25

26

27

28

29

30

1890 1910 1930 1950 1970 1990 2010

year

Air t

empe

ratu

re (C

)Mean summer temperature at NOA (1897-2010)

Trends 1897-2010: + 0.16°C/decade, or ~ 1.8°C

Trends 1976-2010: +0.9°C/decade, or ~ 3°C

Mean winter temperature at NOA (1897-2010)

Trends 1897-2010: +0.02°C/decade, or ~ 0.22°C

Trends 1976-2010: +0.15°C/decade, or ~ 0.5°C

y = 0.0024x + 5.3508

6

7

8

9

10

11

12

13

1890 1910 1930 1950 1970 1990 2010

year

Air

tem

pera

ture

global climate responds to the radiative forcing and feedbacks resulting from emissions of greenhouse gasesregional climates further respond to drivers as land-use change, agriculture, deforestation, and irrigationlocalised micro-climates can develop as a result of immediate human activity and settlements

most apparent and widely known expression of this is the urban heat island (UHI) effect: measurably warmer urban areas than surrounding rural environments

UHI intensity: reflected by temperature differences between urban and rural sites depends on the size, population and industrial development of a city, topography, regional climate and meteorological conditions

Global warming or urbanization?

should be considered on three different scales

First, there is the mesoscale of the whole city

Second is the local scale on the order of the size of a park

Third scale is the microscale of the garden and buildings near the meteorological observing site

Of the three scales the microscale and local-scale effects generally are larger than mesoscale effects

Urban heat island

UHI in Athens is of the order of 4-5°C between urban and suburban stations (Livada et al. 2005)

But UHI can be of the order of 10°C between rural stations and the central zone of the city

Significant research on spatial variability of UHI in Athens, but…

• Studies concerning the rates of changes of UHI over time are missing

• It is important to know whether the intensity of the urban heat island in the city continues to amplify, if it has moderated or has stabilized

Urban heat island in Athens – previous research

Helliniko (HEL) ► urban coastal

Tatoi (TAT) ► rural – suburban

Tanagra (TAN) ► rural (airport)

Aliartos (ALI) r► ural

Elefsis (ELE) ► suburban (industrial)

Selected Stations

NOA represents the ‘background’ urban conditions in Athens on the mesoscale

NOA urban in a park ► (unchanged environment local and microscale)

y = 0.09x - 150.12

y = 0.10x - 168.62

28

29

30

31

32

33

34

35

36

37

1970 1975 1980 1985 1990 1995 2000 2005 2010 2015

HEL maxNOA max

Summer maximum temperature

Similar trends at NOA and HEL

NOA (urban) and HEL (coastal urban) stations

Summer average temperature

22

23

24

2526

27

28

29

30

1975 1980 1985 1990 1995 2000 2005

rura l1rura l2urban1urban2

Summer maximum and minimum tempearure

10

15

20

25

30

35

40

1975 1980 1985 1990 1995 2000 2005

rural1(max) rural1(min) rural2(max) rural2(min)urban1(max) urban1(min) urban2(max) urban2(min)

Summer average, maximum and minimum temperatures at two urban and two rural stations

y = 0.04x - 79.43

y = 0.04x - 79.99

-0.8

-0.4

0

0.4

0.8

1.2

1.6

1975 1980 1985 1990 1995 2000 2005

NOA-TAN (avg)HEL-TAN(avg)

y = 0.02x - 41.93

y = 0.02x - 41.38

-0.5

0

0.5

1

1.5

2

2.5

1975 1980 1985 1990 1995 2000 2005

HEL-TAT(avg)NOA-TAT(avg)

UHI (Turban –Trural)

UHI (Turban – Tsuburban)

Summer UHI Temporal Variation and Trends

y = -0.0008x + 11.586

y = -0.0005x + 11.589

6

7

8

9

10

11

12

13

1975 1980 1985 1990 1995 2000 2005

win

ter t

empe

ratu

re (C

)

HEL avg

NOA avg

Winter temperature at NOA (urban) and HEL (coastal urban) stations

Similar trends (no statistical significant trend)

UHI (Turban –Trural)

y = 0.02x - 37.87

y = 0.02x - 37.87

0

1

2

3

4

1975 1980 1985 1990 1995 2000 2005

NOA-TAN avgHEL-TAN avg

y = 0.01x - 17.95

y = 0.01x - 17.95

0

1

2

3

4

1975 1980 1985 1990 1995 2000 2005

HEL-TAT avgNOA-TAT avg

Winter UHI TrendsUHI (Turban –Trural)

UHI (Turban – Tsuburban)

Summer average air temperature trends after 1976 at two urban stations in Athens are of the order of 1°C/decade

UHI in summer increases at a rate of the order of 0.4°C /decade (urban-rural) or 0.2°C/decade (urban-suburban)

Roughly 60% of warming is due to global warming and the rest 40% is due to urban effect

Winter average air temperature after 1976 at two urban stations in Athens reveal no significant trend. UHI in winter, increases at a rate of the order of 0.2°C/decade (urban-rural) or 0.1oC/decade (urban-suburban).

Results on Athens UHI temporal evolution

Model simulations

A novel climate model that includes a sub-grid urban land-surface model was applied for an integrated impacts assessment for the Athens urban case study

► in order to quantify the cumulative impact and relative importance of

climate change and urban heat island

to the exposure of urban populations to temperature extremes as it is essential to consider the combined role of global warming and local urban warming

T3

T2

T1

T4

H1 H2 H3 H4

Urban Heat Island SimulationsHadRM3 modified to include MOSES 2.2 land-surface scheme and urban parameterisation ►“tiled” surfaceThe "tiles" allow sub-grid variationsNine extracted gridcells over and surrounding the tree cities

1971-1990 2041-2060

NoUrbNoAnth -

UrbNoAnth UrbNoAnthUrbAnth UrbAnth

- Urb3Anth

Urban fraction is set to zero

Fully coupled urban model

Fully coupled urban model including additional anthropogenic term

As UrbAnth, but with tripled anthropogenic heat flux

Surface tile scheme, representing land surface exchanges within a single model gridcell

Urban Heat Island Simulations – gridcells

Gridcells organised as:1 2 3 4 5 6 7 8 9

cell 5 represents the city centre location and 1,2,3,4,6,7,8,9 are surrounding

22 .5 23 23 .5 24 24 .5 2537 .5

38

38 .5

39

1 2 3

4 5 67 8 9

Climate model bias

• Larger RMSE for coastal points than for neighboring inland points, suggesting coastal influence on the apparent model bias

Urban Tile Temperatures - AthensTmin

0

5

10

15

20

25

30

0 2 4 6 8 10 12

NoUNoA

UNoA

UA

NOA

Tmax

0

5

10

15

20

25

30

35

40

0 2 4 6 8 10 12

NoUNoA

UNoA

UA

NOA

Tmin

-5

-4

-3

-2

-1

0

1

2

0 2 4 6 8 10 12

NoUNoA

UNoA

UA

Tmax

-2.5

-2

-1.5

-1

-0.5

0

0.5

1

1.5

2

2.5

3

0 2 4 6 8 10 12

NoUNoA

UNoA

UA

Intra-annual cycle of Tmin and Tmax (upper panels) for NoUrbNoAnth, UrbNoAnth, UrbAnth simulations and observations and the differences (lower panels) between simulations and observations, using urban tile temperatures, between 1971-1990.

RMSE1.501.541.49

RMSE2.442.091.74

Gridcell mean temperatures - Athens

Intra-annual cycle of Tmin and Tmax (upper panels) for NoUrbNoAnth, UrbNoAnth, UrbAnth simulations and observations and the differences (lower panels) between simulations and observations, using gridcell mean temperatures, between 1971-1990.

-6.00

-5.00

-4.00

-3.00

-2.00

-1.00

0.00

1.00

2.00

3.00

0 2 4 6 8 10 12

NoUrbNoAnth

UrbNoAnth

UrbAnth-3.00

-2.00

-1.00

0.00

1.00

2.00

3.00

4.00

0 2 4 6 8 10 12

NoUrbNoAnth

UrbNoAnth

UrbAnth

Tmin

0.00

5.00

10.00

15.00

20.00

25.00

0 2 4 6 8 10 12

NoUrbNoAnth

UrbNoAnth

UrbAnth

NOA

Tmax

0.00

5.00

10.00

15.00

20.00

25.00

30.00

35.00

40.00

0 2 4 6 8 10 12

NoUrbNoAnth

UrbNoAnth

UrbAnth

NOA

RMSE1.481.511.39

RMSE4.342.742.42

RMSE1.481.511.39

RMSE4.342.742.42

Urban model bias – Athens case

greatest impact of anthropogenic waste heat on the UHI during winter

during summer, it becomes a small term relative to the solar forcing

gridcell-5

0

0.5

1

1.5

2

2.5

3

0 2 4 6 8 10 12

UrbAnth-NoUrbNoAnth

UrbNoAnth-NoUrbNoAthn

Model evaluation

22 .5 23 23 .5 24 24 .5 2537 .5

38

38 .5

39

1 2 3

4 5 67 8 9

Summer Daytime/Nocturnal UHI UHIurban-rural

obs NoUrbNoAnth RMSEdaytime 0.16 1.14 0.98

nocturnal 4.55 2.72 1.83obs UrbNoAnth RMSE

daytime 0.16 1.06 0.9nocturnal 4.55 4.57 0.02

obs UrbAnth RMSEdaytime 0.16 1.07 0.91

nocturnal 4.55 4.75 0.2

UHIurban-ruralobs NoUrbNoAnth RMSE

daytime 0.55 0.71 0.16nocturnal 4.39 2.00 2.39

obs UrbNoAnth RMSEdaytime 0.55 0.71 0.16

nocturnal 4.39 3.93 0.46obs UrbAnth RMSE

daytime 0.55 0.77 0.22nocturnal 4.39 4.11 0.28

UHIurban-suburban

obs NoUrbNoAnth RMSEdaytime 0.59 1.07 0.48

nocturnal 3.77 0.45 3.32obs UrbNoAnth RMSE

daytime 0.59 1.16 0.57nocturnal 3.77 2.21 1.56

obs UrbAnth RMSEdaytime 0.59 1.13 0.54

nocturnal 3.77 2.39 1.38

Winter Daytime/Nocturnal UHI UHIurban-rural

UHIurban-rural

UHIurban-suburban

obs NoUrbNoAnth RMSEdaytime 1.50 1.38 0.12

nocturnal 3.50 -0.07 3.57obs UrbNoAnth RMSE

daytime 1.50 1.44 0.06nocturnal 3.50 1.69 1.81

obs UrbAnth RMSEdaytime 1.50 1.51 0.01

nocturnal 3.50 1.99 1.51

obs NoUrbNoAnth RMSEdaytime 1.20 0.77 0.43

nocturnal 3.40 -1.09 4.49obs UrbNoAnth RMSE

daytime 1.20 0.81 0.39nocturnal 3.40 0.68 2.72

obs UrbAnth RMSEdaytime 1.20 0.87 0.33

nocturnal 3.40 1.04 2.36

obs NoUrbNoAnth RMSEdaytime 1.40 0.38 1.02

nocturnal 3.00 0.67 2.33obs UrbNoAnth RMSE

daytime 1.40 0.45 0.95nocturnal 3.00 1.94 1.06

obs UrbAnth RMSEdaytime 1.40 0.5 0.90

nocturnal 3.00 2.13 0.87

Future Projections of UHI for summerUHIurban-rural

UHIurban-rural

UHIurban-suburban

UrbNoAnth %daytime 0.97 -8.5

nocturnal 4.67 2.2UrbAnth

daytime 1.05 -1.9nocturnal 4.66 -1.9

Urb3Anthdaytime 1.12 4.7

nocturnal 5.16 11.0

UrbNoAnth %daytime 0.83 16.9

nocturnal 4.40 12.0UrbAnth

daytime 0.75 -2.6nocturnal 4.18 1.7

Urb3Anthdaytime 0.86 16.2

nocturnal 4.69 16.7

UrbNoAnth %daytime 1.14 -1.7

nocturnal 2.17 -1.8UrbAnth

daytime 1.17 3.5nocturnal 2.29 -4.2

Urb3Anthdaytime 1.27 11.4

nocturnal 2.72 18.3

Future Projections of UHI for winterUHIurban-rural

UHIurban-rural

UHIurban-suburban

UrbNoAnth %daytime 0.97 -8.5

nocturnal 4.67 2.2UrbAnth

daytime 1.05 -1.9nocturnal 4.66 -1.9

Urb3Anthdaytime 1.12 4.7

nocturnal 5.16 11.0

UrbNoAnth %daytime 1.09 34.6

nocturnal 1.26 85.3UrbAnth

daytime 0.88 1.1nocturnal 0.95 -8.7

Urb3Anthdaytime 1.06 26.2

nocturnal 1.52 90.0

UrbNoAnth %daytime 0.41 -8.9

nocturnal 1.99 2.6UrbAnth

daytime 0.52 4.0nocturnal 2.15 0.9

Urb3Anthdaytime 0.65 38.3

nocturnal 2.46 21.8

Model Results

The analysis of the influence of the urban land surface and urban anthropogenic heat emissions for Athens showed that:the main characteristics of a Mediterranean UHI captured by a simple urban surface exchange scheme, when compared observational data for Athenstemperature changes in response to an SRES A1B scenario by the 2050s are similar for urban and nonurban surfaces future climate change will affect the urban and rural areas as wellfuture UHI responds to changes in the additional driver of anthropogenic heat emissions of a city

Thank you for your attention!

The model simulations were performed within the context of EU-FP6 project CIRCE Integrated Project-Climate Change and Impact Research: the Mediterranean Environment (http://www.circeproject.eu).