Transcript geog310_humidity

Urban Moisture, Cloud, and
Precipitation
GEOG 310 Urban Climatology
Dr. Hengchun Ye
1. Difference of Absolute Humidity (gram of water vapor/total kg of
air) between urban and rural differences rather small and spatial
pattern often complex:
a. Urban canopy air usually is drier by day, but more moist by night.
(During the day, greater rural evapotranspiration (ET); starts early
evening, rural air cools faster, ET decreases, and moisture in lower
layer is depleted by dew formation throughout night. Therefore, rural
humidity decreases through the night, a vapor inversion forms and.
(In the city there is less ET, reduced dewfall, anthropogenic water
vapor input and stagnation of airflow, which all combine to maintain
more humid atmosphere)
(sunrise-evaporation of dewfall and other surfaces, water rapidly
replenishes rural moisture as convection slow to develop)
b. Nighttime moisture island in the city is similar to temperature. In ideal
conditions, humidity island also has a ‘cliff’ at urban/rural boundary.
During the day, city drier. Winds can decrease urban/rural
differences.
2. In cold climate regions, city in winter has more humidity
during day. (Rural source of vapor, ET, is eliminated and
vegetation dormant, but city’s anthropogenic releases
from combustion, especially space heating, provide
significant vapor input).
3. Relative humidity (RH=total water vapor/holding
capacity at the given air temp). City’s annual mean is 6%
less, winter is 2% less, summer is 8% less than rural. RH
is a function of temperature (the holding capacity
increases with air temperature; due to heat islands, RH
is smaller).
4. Fog. City is more foggy (on average winter has 100% more and
summer has 30% more fog in cities).
a. Problem of definition, based on visibility to measure fog, which is
always less in cities and makes no distinction between pollution or
water droplets.
b. Dense fog, visibility less than 400 m, often less in cities than suburbs
or rural areas. Improvement may be due to both heat island and
abundance of condensation nuclei in city. (Increased number of
nuclei results in greater competition for vapor and larger number of
small droplets, which don’t produce dense fog)
c. However, at high latitudes, urbanization leads to ice fogs. (Release of
vapor in -30°C or less causes fog of ice crystals as saturation vapor
pressure very low. Combustion of fuel for heating, industry, aircraft
and especially cars, mainly responsible)
5. Clouds. City averages 5-10% more.
a. Clouds are influenced both by increased convection and by
enormous production of hygroscopic condensation nuclei
(produced by catalytic oxidation of sulfur dioxide and hygroscopic
sulfates, oxides of nitrogen transformed to nitric acid; many other
small particles). The former is more likely to produce clouds in
summer, the latter more apt to cause early condensation at inversion
layers in higher winter humidity.
b. Summer cumulus clouds form earlier over cities than rural.
c. Anomalous cloudiness over industrial plants and cooling towers.
(Example: St. Louis, first cumulus formed 3 times expected
frequencies over central urban area and refineries to N. of city,
which are sources of heat, vapor and nuclei. Cloud condensation
nuclei (CCN) increased 54% at low flight levels from upwind to
downwind of St. L. Also average droplets diameters are smaller
downwind by about 2-3 μ (microns) and more uniformly small due to
greater competition for vapor.)
6. Precipitation is the most puzzling of urban
meteorological variables. It is complicated by topography
and water bodies, which stabilizes air masses in
summer, but furnish moisture in winter.
Table 8.7 (handout)
urban-rural
differences of annual
precipitation reported
in the literature
(urban climate)
Precipitation-continue
a. There is much evidence for increased precipitation in urban areas over
rural areas (downwind).
b. On average, cities are 5-10% more in precipitation amounts; days with
0.2”+ have 10% more in cities.
c. 3 main contributing factors for modification and augmentation of
precipitation:
i. heat island causes rising motion over cities.
ii. obstacle effect-aerodynamic roughness of cities impedes progress of
weather systems. Slower movement leads to increased precipitation than
when moving fast.
iii. pollution-CCN can promote or inhibit precipitation. For ice nuclei,
particles might initiate precipitation process in super-cooled clouds. Some
evidence shows that pollution effect is secondary to aerodynamic and
thermodynamic impacts.
Maximum urban-rural differences in summer
rainfall and severe weather events (handout)
Percent increase in thunderstorm incidences as
related to urban population size (handout)
d. Network approach. Setting up large network of instruments show
upwind vs. downwind differences. Need to make sure there are no
upwind controls, such as other urban, topography, and water effects.
• Ex.1 St. Louis upwind-downwind differences 23%, statistically
significant.
• Ex.2 Tel-Aviv, Israel. With 3 decades of record, shows 5-17%
increased rain downwind.
• Ex.3 Turin, Italy. Between 1952 and 1969, city grew 700,000 to 1.2
million and cars from 70,000 to 390,000. Summer shower events
increased in frequency in the city, but also decreased in
volume/shower. In winter, there was also influence from the air of
Milan, an industrial city. There was a decrease in both volumes and
frequency of precipitation in Turin. But adjacent rural zone had a
relative increase in events ascribed to nucleation by industrial
aerosols.
• Ex.4 In W. Europe, colder season showed increased days with
drizzle in industrial cities.
e. Refineries produce sulfates and nitrates. Nitrates are
suspected of being more active nuclei, being larger and
more hygroscopic than sulfates. Their presence leads to
a wide range of cloud drop sizes, which are better for
rain. Sulfates are small particles and more apt to
stabilize clouds-possibly more fog. This shows that the
nature of aerosols are important, not just their numbers.
f. Weekdays vs. weekends. Paris showed 24% decrease
on weekends. 22 E. US cities also noted weekly cycle,
but only in winter, with weekends 8% less. Rainfall
pattern followed pollution levels. Recently, some cities
showed more pollution and rainfall on week days.
g. Slow-moving storms show heavier rains over urban
areas. Several cases of floods in cities while rural
surroundings had considerably less.
Examples: 1. Cloudburst over Berlin on 6/27/64 with little vertical motion, a
cumulonimbus developed over the city leading in 30 minutes to electrical
discharges, rain and hail. In 2-hrs, 81 mm fell while to the SE and NE almost no
rain observed. 2. In Nov., 2004, a storm cloud stalled over central L.A. dumping
over 5” rain, lightning and several inches of hail, while very little rain fell elsewhere
in the urban area.
h. Both summer rain amounts and thunderstorm frequencies seem to be larger just
downwind from cities, but not all over city. (In Berlin, Germany, average
thunderstorm days number 32 in a SW suburb, only 20 in the inner city and 28 to
the NE. Summer maximum rainfall in SW approaching zone for storms has 200
mm, inner city 160 mm, with 190 to NE suburb. No explanation was made for
avoidance of inner city by thunderstorms.)
i. Fujita noted curious avoidance of central Chicago and central Tokyo by tornadoes.
Horseshoe area located over Chicago appears tornado-free during past 20 yrs.
Also there’s a horseshoe-shaped tornado-free area around Tokyo. The opinion of
Fujita is that heat island so uniform that significant vorticity (spin) mainly generated
at interface with cooler country air. Also the urban friction assumed to decrease
tornado intensity.
j. Changnon showed there is an increase in thunderstorms and hail for cities over
rural areas and that the increase is proportional to the population size of the city.
(Ex. Laporte, Ind. Anomaly concerned a 100-200% increase in thunderstorms and hail
over a small town 35 miles SE of the Chicago-Gary, Ind. Industrial source).
k. Snow. There is less snow in cities (heat island) as more falls as rain (in Lund,
Sweden, snow depth decreased from 8cm in rural area to 3cm in urban center).
But another effect is in cold climate, anthropogenic nuclei (cola-fired power plant
produce ice nuclei, some of which are already effective at -5C, can cause snow fall
related to the plume) induce more winter snows.
Urban Wind Field
The changes brought about by urbanization in the
local atmospheric boundary layer have a notable
effect on the low-level wind. This is caused by
the Heat Island and the change in surface
roughness.
Surface Wind
1. Fig 6.1 change in wind speed classes
during rapid growth area of Columbia,
MD.
Three wind categories, (a) Columbia wind
speed was less than 70% of that at
airport; (b) Wind speed is 70-99% of
airport; (c) wind speed is higher than
airport. The frequency of (c) was 25%
in 1969 and dropped to 14% in 1974.
The frequency of (a) decreases from
43% to 65%.
Exampes: Wind speed decreases in
Gantsevitchi, Russia from 3.9m/s to
2.5m/s. Also in Partma, Italy.
2. Wind direction of greater speed is less
affected by urbanization than weak
ones.
3. When regional winds are light and skies are clear, a drift or
convergence of surface air towards the urban heat island and a
compensating outflow aloft have been recorded empirically in different
cities. (This is a logical consequence of the heat island, which creates
an unstable vertical lapse rate of temperature and so induces a rising
air current).
4. Centripetal urban air flow (into the city) is usually more pronounced
at night than during the day because the temperature difference
(urban-rural) is greater at night.
5. When air flows from country to city, it encounters a great
increase in surface roughness. The wind near the ground
therefore slows down except for local accelerations
around individual buildings, and becomes more turbulent.
Landsberg (1981) states that there is a reduction in
speed of 20 to 50% at the standard observing level of 10
m.
Changes in Mean Wind Speeds in Parma, Italy in Three
Consecutive Decades*
Interval January April
July
October Year
1938-1949
0.5
1.8
1.8
1.0
1950-1961
0.5
1.4
1.4
0.7
1962-1973
0.3
1.0
1.3
0.6
*wind speed (m sec-1); adopted from Zanella (1976).
1.3
1.0
0.8
6. When there are deciduous trees in the city, the reduction in wind speed
increases in summer and decreases in winter.
7. A differential day-night wind speed reduction has also been observed in
different cities. Lee (1979) clearly showed that if in the daytime the wind
speeds in London were reduced on an average of 30 percent, the average
nighttime reduction was only 20 percent. In some cases, a strong nocturnal
convergence results in greater wind speed in the city than in its surrounding.
8. The flow of air around buildings and in the street canyon
is very complex and depends greatly on the design of
the city.
Vertical Wind
1. One of the common mathematical models describing airflow over a city was
proposed by Davenport in 1968:
•
U*/VG = (Z/ZG)α,
•
where:
•
U* is the wind at height Z
•
VG is the gradient wind speed at height ZG
•
α is dependent on surface roughness
Measurements of the vertical Wind Profile for Different Surface Roughnesses*
• Terrain characteristics Exponent of power law
Gradient wind level (m)
• Open country, flat
0.16
270
• Suburban settlement
0.28
390
• Inner cities
0.40
420
•
*Davenport (1965).
2. This formula is good for determining the average wind over a city
(but does not help in predicting and calculating turbulence and peak
gust winds. Generally the wind speed increases with height.
However city friction reduces the rate of increase with height,
reaching the non-friction level at a much higher level -gradient wind
level).
3. Urban areas have also an affect on the wind direction and
turbulence. There may be more turbulence within the urban
boundary layer than in the surrounding areas.
4. Narrow streets between tall buildings can have a funneling effect, in
which wind is actually speeded up when forced to flow through a
narrowing or bottleneck.
5. Edges of buildings and the geometry and spacing of buildings may
cause turbulent eddies (small whorls) as wind flows around these
obstacles.
C. Practical Importance of Knowing the Wind Structure
1. Air pollution control: the important variable is the
concentration of pollutants at ground level,
which is a function of wind speed, direction and
turbulence.
2. Building heating and ventilation.
3. Wind loading on structures (wind stress).
4. Windbreaks to minimize drifting and blowing
snow on expressways and around buildings, in
cold climates.
5. Coastal cities prone to damaging winds, such as
hurricanes.
D. Conclusions
The urban wind field is rarely simple. Even when
the synoptic weather situation is least
complicated, small differences in local
topography will cause irregular air flows.
River valleys, bays, and shore locations on lakes
and oceans develop their own secondary
circulations, such as land and sea breezes, and
mountain and valley breezes. (They interact with
the urban-induced circulation and may, at times,
dominate it).
• Land/sea breeze (coastal cities)
06_18.JPG
Mountain/valley breeze (mountainous cities)
06_19.JPG