lu.bmttte.ct to Journal of Energy Efficiency, Vol. t, No. l, 1993.

Mitigation of Urban Heat Islands: Materials, Utility Programs, Updateat

Arthur H. Rosenfeld Hashem Akbari

Sarah Bretz David Sailor Haider Taha

Heat Island Group Lawrence Berkeley Laboratory

Berkeley, California 94720

Draft: May 19, 1993

Abstract

Elevated temperatures in urban "heat islands" result in high cooling energy use and may enhance the formation of urban smog. Urban shade trees and light­colored surfaces can lower the air temperature of a heat island and reduce cooling energy use. Light-colored surfaces stay cool because they typically have a high "albedo" and reflect solar radiation that would otherwise heat the surface.

The albedo of a city maY. be increased gradually if high-albedo surfaces are chosen to replace darker materials at the time of routine maintenance. Most paints and roofing materials are available in light colors at no additional cost. Pavement albedo may be increased by using concrete or by using white aggregate in asphalt pavement and rolling on top an inexpensive layer of chippings, shells or sand. Light-colored surfaces may last longer than conventional dark surfaces because they reflect damaging radiation, stay cooler, and suffer less thermal expansion and contraction.

Using white surfaces to increase the albedo of a city may be a lucrative way to conserve energy and reduce pollution. Utilities could promote high-albedo surfaces

t An earlier version of Sections 1-5 of this paper was presented to the NIGEC (National Institute for Global Environmental Change) "Supercities Conference," San Francisco, October 28, 1992, and will be published in Urbarr Atmosphere/Atmospheric Environment (1993).

to earn profits and:reduce the demand for peak power. Suggested progr.ams include 'paint g.labeling, ,marketing., and incentives· 'ifor using light-colored ,i,na teria..ls-: for buildings.and roads. ~-'! : "'.r ::. · ~ "( ; ·:1d · :

• 1 '· · c ; ·_/ ·,; ' .'I' ·;; ,_., i;.:rt: (); ..

. ~:- r· · W..e: ;present rtMee recent meastlrements: the air-celnditiontng '· savings on a test house in Sacramento, on two houses in Florida, and some ·-:ail'l'·rtemperature measurements at White Sands, New Mexico. In addition, we discuss the results of some meteorological simulations performed for Los Angeles, CA, ·!tlhat-rv.alidate the

;~~:~ure~ ~~ta fr~~>l~~~ite Sands:J ·~:, 'l )i;<,s ,·;;:

· ·1: ; ·Introd1.1ift1"·on .·, · ·· 1 :... rr., · .... ;:i:; ;;.~ .. :::: ·-· ,..,....,, ' .3 ... \ • ~~

;, · ' .,,:; 1:· , ., ... ;.i:. di ' :'. ilO(>'.'

,: ~ . On a summer day, fhe average tem~rature in a typical American city is::alnlixut SJ F hotte~ than-the surrounding rural area. <JDark surfaces:. tl\at heat up as, they absorb solatrr~iation~ ·and the reduced vegetation,~contribute to what is termed th~"urban heab<island." Akbari et ·al. (1990) estimate that ,.~·a:o% of urbannpeak electric demand toJ.iay is for additio~ ·air conditioning to ,.compensate'".fbr, 'he.at!:tislands, costing ratepayers over an additional· $1 billion per. ~ar.:. pluS.1 '.th:~,(·Iarger ·indirect cost of pollution. Elevated temperatures enhance the formation of smog and urban ozone, the pollutant that most ofte.n,:exceeds.the N'ajtion~l: Ambient Air Quality Standards.

!s;: . . '1P: ::r:• . ;.on; · · ·:,· ,, .. :.•. ,.;,:;,t:, . _

.· ·-~ '..: Trees and · light~colored :surfaGes are inexpensive and effective ways to mitigate urban heat islands. Light-colored surfaces typically have ·a high "albedo," meaning that they reflect a large percentage of the sun's radiation. Because they are good reflectors, they stay co.Oler than absorbing stllfaces. Lower surfacetemperatures reduce,the city'~jcooling· energy use by directly '< reducing the heat gain throu!}h ·a !building's en\r.elope (dire.ct effect) and. by lowering the air temperature;. (indirect 'effect). t - ... ;. \, " - , .\:' ~i1t~ ::. < r•·= •. :.·~ ' : ;.'

~'. HeaMsland Effects:·;-,.' " '(

Summer urban heat islands with temperatures 5-8°F higher than surrounding areas are 1ouna throughdur the U1S. In fange cities like 'tos Angeles, summer '·monthly averag~ tem:pe~-atur.es are increasing -f aster tha-l\ :1°F per decade (Fisuter. 1). ! his ~warming .. -rais.es peak caoHrtg defnand 1.5% (Los . Angeles) to 3.0% (Atlanta) for each 1°F risei On a hot afternoon, U.S. heat .fsla'l\ds· raise air conditioning demand by about 10 GW (gigawatt), worth several million dollars per hour, and integrating to several billion dollars annually (Akbari et al. 1990; Competitek 1992).

~ f ...

Heat isfands have several effects on urban air quality. The power needed to compensate for heat islands requires significant additional · genet'ating :i capacity, which contributes to urban air pollution. This contribution .ma~ bf! significant, since peak power is often supplied by t~ most polluting power ip:lants: Fllrther, elevated temperatures associated with heatJislands accelerate the formati~n. of.smog. Figure 2 shows that the probability of ·smog increases by 2%· to 4%tlpei' ?F: .• Belo'.INI 70°F there

, "\It • •" • I • ~· ,'f •• /uJer3/bol/h_1h1g/CIEE/1993i(:IEE.PAP/tu~

.-3-

are almost ll\eve~ smog ''episodes" in Los Angeles, but starting at. about 73°F, smog episodes.> begin and:;exceed 50% by 90°R ·We estimate ~rudely :that reducing ,daily highs by 7°F (i.e., just by eliminating the heat island, but we C.Ould d~·bet1;er:_othan that) will eliminate an astounding 2/3 of the smog episodes (Akbari et al. 1990). As discus~edvlate~)n·this paper, we have started meteorological and smog mo'deling to refine-this mude·estimate. ::::•:~: · .,.·

~). /'. :-;;, _" ~' ' '~l

3. Treea.andcWhite Surfaces i ''I .. ,.

. ' j\.'

Shade trees and high-albedo surfaces are effective ways to mitigate urban heat islands. Through direct shading and evapotranspiration, trees redhaet-<sum,!l:nei' cooling energy use in buildings at only about 1 % of the capital cost of avoided power .pial\ts and air-conditioni•ng equipment i(Akbari et al.,, 1,990). Light.i:colored surfaces should be even more . effective than titees at cooling Cities, and cost lessl'if coltir changes,~are incorporated into routine maintenance schedules. Also, tH., results from light-colored surfaces are immediate, while it may be ten or more years~before a tree is large:ie~ough:tb !:produce·rsignificant energy sav.ings; Akbari et al. ·(19.89) discusses the relative.benefitlrcost!of.wll\ite surfaces vs .• trees.'"; .... , .:>';.:·· ·

.~. . .... ' t ~.·' ~

High-albedp materials om,) majQr: urban surfaces: such' as roofs, streets, sidewalks and parking lots, will be cooler than the conventional dark materials. Figure 3 demonstrates the r.elationship betw,een, ;albedo and" surface· temperature of various materials. i" L :l ',:'.'.' · ·: · · ';' ·

. ·,:· .

. " ....

. · .; ';; The .:effects of light-colored. surfaces include 11direct" energy savings and "ingirect" effects on climate. Direct effects refer to. 'th_e energy savings of an ;individual building achieved by .Changing the albedo of .its~roof. If albedo is altered on a larger, community scale so that the ambient temperature decreases, the energy savings of buildings in the area are indirect effects of albedo modification. Indirect savings are achieved because in a cooler environment, it requires .less energyuto -cool a building.

. ;.'.: ' .' ~ ' J '

Computer simulation fit to. a ··;1real house in .Sacramento has .shown .that applying ~ white coating '.to. the ·roof, and adding.:thr,ee sh:ade tre~s wiU•!teduce the cooling elect).;;id'ty;use by nearly 5(:)% ;(figure 4). Additional:savings~can be·achitved by using high-albedo materials on major urban surfaces.· il/; '''

t : I •I · ~ ,'i 1: . • ,

4. Costs . ,, ·. ~ ' ... f ~ ; I ~' ' 'i

Increasing the albedo of a city may be very inexpensive, if the change is made at the time ·of ~utine maintenance. Buildings are typically repainted every 10 years, and whit'- :paint<·~y be used for no additional cost. Similarly, many roof types, including single•¥f:f! .. membrane and built-up roofing, are available in white for no additional 1cosf.. For p~vement, light-colored pavement ··could be installed at the

. time.~of r.esurfadngb,: r "·Whitetopping" (resurfacing an asphalt pavement with concret~} produces: a "light-colored pavement that has low maintenance costs and a

j ~· :t · .. ,. . \ • .. ,' .. /u1er3/bea/h_ahag/CIEE/1993/CIEE.PAP/1uper6

·1-4-

long service life. Another option for light-colored pavement is1 a "het-roll~d" asphalt pavement using white aggregate. This paving method, popular in Great Britain; initolv~s rolling chippings intd-.the t0p surface of the pli~meiR· 11f white

··chippifigs a~ 4oc:ally availa~le, a ·;Hgll.t;;colored- pavement is ! proC'!lutetl at"'~rio · additiorial cost .. oWhite chippitl.gg can al~ be used in a "chip seal" surface treatment.

· \~ ... c :-.. dL ~- -~ (t _,.J ~=f:~·i· . ;?

°' ' . It appe·arS·'.·that most ,:Ught-cotored surfaces ·a·re eq1fr(lalent· Jn cosf'ifo I conV:.entional .dark surfaces and last longer ~taus~th'ey are -less susceJ'Hble ~to '. the ··damaging effects of solar radiatitin-and large temperaf.ure swihgs!.- 'While no ' quantitative aata have been· found to1 support this hypothesis', 'it i\:S f'Corrim0n knowledge among roofing professionals that white asphalt sltihg\es last t6ngerltfilan dar.lcasphalt shingles. Since "white" asphalt shingles are in ~ fac-Pquite dark' i'.(~% absorptive, see Fig. 3) when compared to white paint, it would •seem.r:that small improvements in albedo can significantly effect service life.

~ .J A :~· L : 1J: ') : ! ~ .1 ::3;"":·:;)1:

S • .l.a·beling and lncent_ives for Cool Surface~ .'1:''..· ' . . . ~; f. r· ,,

c .• ,.. ,, S.· f ~ : ~~. ':j· .

. :. ·. W~ estimat~· · that heat island redtlctiori. savings~ of several bilHon dollars , annually could" 'be ·r.ealized. through-·utility,;speh-sored,"~demar{d-side managemeht (DSM) programs that promote the whitening and greening of cities. Assuming these utilities are permitted to retain 10% of program savings, they could earn about $100-$300 million/year (Rosenfeld et al. 1992).

Sa. Labels- . -" ' . -

Some utilities ·haye already expressed interest in incentives, :i b~t..:.lfucentives for what?. The problem .isthat we want:cool'.surfaces, and Fig. 3 immediately shows that ~here is little correlation·.betwe~iv'.'fOOl'' and "light-colored.'!;·" J\n incentive for the "white" asphalt shingle at 175°F,rot1 ~aJ,vanized steel at 170°, would not only be counterproductive, but would miss the opportunity of making a huge ,new :market for c~oli, lj~ht shingle, coo~-surfaced ;galv~mzed steel, and smart pai~ts ~nd surfaces (pleasanthght colors, which stay cool lbeca'l,lse they are very reflective-dn the near­infrared)~, Thus, labeling is an essenti,alJirst step.

. • i.~ ~ ' . .

Labels can be designed after we hold workshops for the pa·int/pigment industry, the roofing inpustry, and _the pavement industry. These lab~ls will probab~f'. carry the folloWing informatiqn: ...• :F~1··f'

1. Surface temperature on a clear day under the midsummer sun in, say, Arizona. ., ; ,; . ~ . ·'! /.' ,

2. Surface t~qiperature under standard fire co11ditipns , (~o all~w b~tter protectfqn of roofS'and wall~}rom external fires). <> ' L ~·,:

3. Service life (to give credit for the fact that cool roofs and road~c1ast longer than surfaces exposed to diurnal thermal shock and ultraviolet radiation).

/user3/bCl/h_1h1g/CIEE/1993,CIEE.PAP/lupe:r6

Sb. lnc~ntiveS;: . , • ·"' . ~ f.· . .: . ..

>.. " . '

-5-

1:· . ··'

' . • . , , ;1111-pl~i;neA~ation prog11~ms for . 1 wl\h~ surfaces should _be.: . .designec:tto ~mpha~~z~goof types that cover the larg~st, area in a 1oty. Rough ·~estimate~ bas~d ,on the perce~tage~,share of California's roof:iJlg mar~~ for each ;roof type t (Table ·l) suggest that programs in California should concentrate on the modified bitumen

. '1\d. fibergl~~s a?p11.:Pt shi,l)gle roofs fo~ homes and (>n built-up roofing and modified bjtuµtfP.X\ _rqofl.rg for),commertja_l applications. . »mlb:up roofing can be installed with a ~;ttjte.re~~~th~e coatlng for;,pa additional cosJt while adding a coating to madified b~tµmen ~iµay iP.clUd~ a very Sqtall incremental cost, as mentioned above. ·Since

,c~ating asphalt $):\i~gles is an additional expense not included in installation and .. ~idS{-.the warr_ah!f:yfor the roof, iJ·:will be necessary to induce shingle manufacturers ;Jq. µtals~ :high-all_;,edo shingles. ·: . :s · -,

<·--:7 ~ -~ ·: ;, _; l ./J'

Another barrier to implementation of cool surfaces is the lack of measurements of the indirect effects of 14rge-scale albed<;>4 modification~" Lawrence Berkeley Laboratory is seeking innovative developers to build half of the houses in

. a develop}nent , cp~ventionally a.nd - h~lf with light-coloreQ ' :roofs and roads. Researche~ will th~Jt, dempnstr~te:t( the ·re.Q.µction in air Jemperq.tiir.e ·and s.a-v:ingS: in air conditioning. . ! • , <oii' >. ~i: · · ~tr:"" " ( ;-1 ;,,.

"'

Residential Type Modified·. bitumen Fiberglass asphalt shingle Built-up roofing Wood: shingles I shakes

· •i· " Organic asphalt shingle Metal Cement tiles Single ply Clay tiles Slate . . •; I . '

1"f ·.•

' . . ~. ... • t

Commercial 3 Type

22.4 ·'Built-up roofing:. 'l'S..8 Modified bi tuinen

·-;:i 6.4 )' Asphalt shingles '.'If't£6f,; EPDM . , ... : i .L(j · FVC·::· -;:r '

2.6, M~tal , ;cl.().. ' "Tile·. .

.':::::J.O, .. PGlyurethane• foam 0.3 Hypalon

, ... ··'

3 ~ . 25.4' ":! ' • 7J]..,; !•.

4.4:! ,, 4.4' 'J •

l.: 2:01.~~

F'.:J,1.8·'._,

j0.6··· 0:5 0.4

; '.) ' YCJ.4 0~3· .. '. ~ , '

0.1 Other single-plies •.. -.- 'Liquid-applied

"""T=o-t_a,,...l ---------5=2-.6..,.........,,,T,,...o .... ta"""'I,....--........,....-----= .. -1-:-s4""'!:·7""":4-t. •1 '

' .. ,, .. ,'..,' . ( :·.r~ I J::i ~ ;' ' ''.''I r

Table 1. Areas of Roof Types in California. Rough estimates, based on da~a on · california'~ ·~odfing nfarkef 1Hn dollars) from ' National Roofirig Contractors Association (19?0) and estimates of cost/area from Means (1990), as compiled in BretZ et al., 1992~'. · · ·

J J' . .. 0 -

1

''J!*" o #> • •• I /user3/be.a/h_ahag/CIEE/1993,cIEE.PAP/1upc:r6

-6-

'- ; ~ ·'." :-i (. ·: "! ~ ~~' ~ .J'

' . ./•} ·j :~-; '· f'L~ S ·/.r , - 1~ i:: .

:~ ;. :·. .Because of: the 30-50% savingsnin a/ c bills, incei.ltive programs for light · siii:face51and~ vegetation should work w~U.'but slowly (oecause roofs fast--3Q:-years~and · shade·tiees grow.· slowly). Cool, shaded roads and parking lots are more··ofia <J.tta~dy/.chaUenge: of the commons,. ·and· may require r~tandatGls. Ir{eit-her ca~, 'international .labels will make ·a' market::ofor environmentally benign and. profitable ne. ~ .. :; s rfac .,. r·i~ . '• ' ~- . ,. . I .~ I •.•· \ '• ::: ~ - u es . . ··.. · . . J ... "' ,. • . ·: · . - ~.. : .. ··" \ ·" • · · _.,

New Developments

In the preceding sections, we gave an overview of the causes and impacts of heat islands and mitigation methods. In this section, we summartze soirie new developments in heat island studies. In particular, we briefly discuss the measured

.~ e:ooling energy savings in two(~xperiments (Sacramento" and Flotida studies), the :indirect cooling effects of high:.albe"do surfaces '-ltleasuted· at White Sands, NM, and .same newv simulation effott:S' ·performed for--.Los Angeles which~:validate our measured data ip White Sands; ' .. :_ ... ,, jj

,i6al· Direct·Cooling iri Saci'amento ''1

In· summers of 1'99J :·ra~d: 1992, we condt4eted experim~nts to measure the " impact of white roofs and shade trees on the cooling energy use'i:)f a few buildings in

Sacrament<;>. :· For most sites/we collected data on air-conditioning electricity use, ' tiridoor and outdoor dry bulb temperature and humidity, surface temperature of roof and ceiling, inside and :.o-utside wall te~peratures, solar gain, and wind speed and direction. ' iv~ ; 1 , - ,.• • .• " • . 1 · • i'.

To measure the impact:. of shadeitrees;iwe conducted·i ·"flip..:flop'~ experiment, divided into three periods, using two houses. In the first period, we monitored the cooling:enetgy use of both.~hollses in:.orderc:tb establish a base case relatiQriship (see :Eigur.e ' ,5); )·111 the second and third· periods; eight large artd eight small shade trees were· pla'Cediat one of the sites'. for a period of four weeks and th~n 'were moved to the ·oth'er site.J! The co6Iing entrg)i' l.\$.e of . .the· site without trees indicates what. the cooling:"ener.gy;,,use of .the;shaded site would have been were the,· trees ·.not there. Figure 5 shows savings of 35% from the median a/c load. SeasdnialLanalysiS is underway.

. ,'"'r -~'

To measure( the impa<:t of high~albedo roofs and walls, ,we monitored cooling energy use of';a 'house arid two sChool bungalows. The albedo of the house roof before modification was 0.'1$. ··:rA coat <of-high;oalbedo paint achieved 0.77. A .year later, the albedo of · tl~.eo ditty roof was :0.59, but washing if with soap and water returned the albedo to o.:73. : ! f. c '

·3 f" ! : .: J : t · ;· , ~ ~ ri. · .. ·;. t , ~ r1 .. ~; ~ · ~~ : : :.; ,J,; ~ --: ; "'. -· , -~ -.:' ~-.. i

,.·:· .. .::Daily-.cOOlirig ... energy ·usein kWh p~r day·are'pl0tted ag~inst the daily average ·. drybulb outdoob temperature in' Figure 6 .. '.fhe squares rS?present the cooling-en~rgy

/u.er3/bel/h_aha&fCIEE/1993K:IEE.PAP/lupcrn

-7-

use for the period that the roof was dark (albedo 0.18). The circles represent 1992 data, when the albedo was measured at 0.73. The daily average outdoor temperature that causes .ti\e :air-conditioning unit to com~ on has shifted upward :by 29C. Note that many1~cirdes fall at 0 kWh. The lines ·on:, Figure 6 are regression fits upJo .. 25.0 C.

~. Past. this~point, it ;is difficult to compare ·pre:-roodification and post-.modification data . :l:>ecauseJHh~re a;r.e ,, .no pre-modification data ,,with compara_qle -.. :enviro!lmental .. variables.L,._:rhe- seasonal .. cooling.energy savings at this site are estimated :to:be:040% (330 kWh/yr), but savings during the long, very hot days of summer are only about 30% (2.9 kWh/ day).

- ·:' 1 ~. ; : ~ ·... ·~

Energy savings of 40% were also measured for the school. "·1

· 6b. Direct Cooling in Florida . , ..... jJ

In 1991 and-1992, Parker.· et al. (1993) of the ·Florida Solar Energy~Center {FSEC) conducted experiments and cmeasured the impact.of: ._reflectiv.:e roof ·coatings· on •air conditioning energy userin three ._ homes in ceptral Florida: one~ with a~. w.ell­insulated R-25 attic; a concrete-block-wall house with ·R.,;11 ceiling insulation;eand a concrete-block-wall flat roof house with no insulation. The roof albedo of these houses were modified to 0.73 from initial values of 0.20-0.22. The measured air conditioning energy savings were 10%, 25%, and 43%, respectively. The reported utility-coincident.peak demand between 5 and 6 ~_.p.m. are 28% and 38%i.for R-11 house and flat r.oof house, respectively; -, Figure; ,7_. shows the . roof ait>i'Space temperature and the air conditioning :energy use before and afte~ 1the application:·of reflective coating on. the R-11 house for July 28 to August -3, 19.92 .. ~ -' This study:,has concluded that reflectiv,e coatings-: .are:. particularly appr.opriate for :· existing Florida homes in which the roof structure makes it difficult to retrofit insulation. >-:- _ .

:/~C<f' ~:·I·~ . . t\1); · ~ J F·:.-L .·:.~·~?.L:',·! '.. ,:.;ilf:l:;. _ - ~ , _ ,:~J ... ··. ~ /\ .:.

Figure 1 suggests that if we res~to_red. ·. Los-· Angeles to 1930's" 'CPndi,Pons,:1Jf vegetation · and, road color we might remove the heat island, exceptrfor<abou~:)l °F from cars ancir:the conversion of electricity1

1to, heat. We.:·believe that :we.;could do

, much better based- on the two points on Fig. 3 :labeled "Hypothetical White: . .City," wi.th a· -S,ola:f~ ahso'Itptivity of 50%, and labeled "Average:City,". with an;.absorptivity of 80%,. i.e:;_a· difference of 30%. :', ;: '·

Our group is just beginning to study air temperature at the White Sands .. National Monument, which is a light-colorerl -. strip se-veral miles.,wide, inside the ;-cWhite Sands Missile :&ange in southern Ne.w-MexiC(). The white~;sand is gypsum, :;which has the doubly interesting .properties of being quite white .(the dunes have an albedo of: 0.65, so an. ~bsorpti:vity of 0.65), and being:;veqt; alkaline, so that even between the dunes the vegetation is sparse. Near the middle of: the Monument,, the average absorptivity is about 0.4. The surrounding desert, at the same altitude of .about 2000 ft., is veg~tated : wi.~h .. yellow-green bushes ·with an albedo .of about 0.2 to <J;4; .so an absorptivity of about 0°.6 to !>.8. Hence, the absorptivity difference between

·• /user3/bca/h_1h1g/CIEE/19931CIEE.PAP/supero

-8-

the monument and the desert is about 0.4, or comparable ,with the conc~ivable improvemerit ·in the albedo of large fractions of a city like Los Angeles (but not dense high-rise downtown areas, like parts of Manhattan).

1 ~ • •• As _~rr~Wn in Figure 7.1, the summer hig~ t,~'mperatures in t~e ~~~gc;?_tated ·d:s.th't ate typically 105~F. Although the d{J~f -.r~. ·very preliminaIY,, :-~~~-~~l\d it 1 encouta~ing that the liSht·colored monumetl.t riins"'about 10°F cooler. · W.e.,chose to

"'I .. • .....,, • j .. ! ~ ..,.1 '

'CJ.tiOte morl'thly highs for the summer months, b~fiause ~hese are jus,t ,the . sort of ·,~o~~ffio1~~ ct~der which urban h~~t islands ~re the ;~orst. . ·· : ; ,1 ••

• ~ •..,:I, • _ _; . ,. ,~·- ;, • - ,. .c... r} I ' t • ~ f.i ~j ·.~, ' . l· • r ~.: '"j. ''r~·r, 'The data of Fig .. :~, .syggesl'the follo\ving ' r~Iationship fo,r the. ; .4.~J?.ress!~Q.vit?f

.:s~nun~~~' p~~k temperatttt~)?l a city-sized, .:1,l~~.:_albeqo are~;, :: . 1'

_ j :-

. 1 ,. '. ' . ~

( 'r-+-( (1) '• ., ,_ .

. 1-..

. :where 6i;, is the c~ange in tempera,ture . (rpeas~red i~! ~el¥!ps, K, where 1 K = 1°C) ~~ .1a is the change in albedo. We: are ,.er~poll!ftged thp.t ·P.t~teorologi~al m~deling of 5~~si'ng the ·albedo of ~~~ .~os. A~~~l~~_,.'6~?h\:Broduced _simil~r results: -~ f; :; •

-. -~. '\ ''• !"" . " ;: '.J " . ' "-:"'"

.1 t · ' . 5.4°F ·= 3K .. ' I< ::li·--= 016 = 19 a-

.)~ •... r"r r:

;:. I •

:: 4.J ... . _;-

..> -·'· (2) ;_

(S.ee Section 8). ,. ' . ',. · . r .. ·~.:.

' 1' ~ . -

Since we estimate that of the present 7°F . summer heat island in Los Angeles, 3-4°F comes from dark roads and roofs, it then looks easy to reverse that 3-4°F, and we could actually aim for another 5-7°F of cooling. Figure 2 show~ JP.~;pqt~~ti.a1 gain in power and smog from cooling in Los Angeles 10°F below the present su~!ll~f.,~O?ditions: about3_ GY:J ~XYHe.ak ;powet-J\Yo;th $?QO,OOO/hour) ~nq a 20% r~~PSUR~·ir:'J~e prob~pility of_ a : ~~og ~pi~o,d;e!c , , . \:::: ,.. ~.,

,•. "f : · . . ; .. _,._ • . ·- ~ ·, -~ ;' .. ~ . N ._.~) ~ d ' :• ' \:. ; '. ~ .... i · , c . ~_Nq'W that w~)1~y,e iritrodii'.c~~ . ~,he ~onc.ept of ~a Jp°F ~rOp· in t~!Jlperature for a

353' ·fncrease in whiteness we can also address t~e alternat~ve Hyp>_~,!h~ti~:al_.,Green -City of Fig. 3, which is 0.1 darker but has enough trees to cC>Ol by evapotranspiration (EJ;) and shacU_ng . . we estimate that E1:; and~~¥de wil~ COQl rthe city by 2-3C?f, which will j~st about' offs~~ th~!t_3~4°F from th~_'incr~se~ absorptiv!,ty; still leaving us with a potentially nicely coolep. city. . . _ Li . . · ..

,, ~ . . - ~

8. l\1e~eorological ~o'd~iing of Albedo it\ ;L.os Angeles, :· : . : ;. :: _; .:,

. - < .L

. . - . .: . - . . t: ' . -- : ·. ,: ·. --~:- . .. ~

.In·'' '.~~other : . ~-t~B r.~.,~~, .u¥ a,~ three-dimen~ional mesoscal~·., qte~eo~qlogical model to study the impacts of proposed surface modifications in the Lo~ Angeles Basin (Sailor and Kessler 1993). The grid for these simulations covers the entire Los Angeles basin extending 325 km east-west and 200 km north-south. Each of the 2600

/user3/be.l/h_1h1g/CIEE/19931CIEE .PA P/supero I • ~ -~

-9-

... ,c_

surface grid cells is -'5 km on a side. A land-use data base was used to.determine the breakdown of land.;;.use in each grid cell. · ,, · ,1 . ~, _.

In d~yeloping the h~gh albedo modification scenario, 394 g,:~d cells were iden'tifi'e-Cl '·as· "developed are·~,. ~Uitable for modification. In evaluating the level of

·" pote~~~~~'."'~lb~d? i~crease we ass\i.~~~ that typical "~e~elol?~d areas" :¥~~~~~"' tp1an half cov~.red with rmpermeable ... ~urfac~s, and that existing fechnology q;>uld, ~n~q;ease the albedo of any individual ~~perme-able surf~ce by 0.30 to 0.50. )t . i,s, th,e,~.fore reasonable that the albedo of an entire 'city could be increased by 0.15 to 0.25. Figure Sa illustrates the loi;:ations of aJ~edp. ip.odificatio~ invest~gated in this. study. The

1avera~e alb'~do increase in each,. ~f the ~·94 modified gri4r c~l~s for this shf,(jy ~as o~~~· Note that an albedo increase of 0.16 'd0es not imply a 'brilliant white city; instead it corresponds to a Los Angeles which is the color of weathered concrete. If, however, white cities become popular (as currently in: Ari:ZOna) we' eould perhaps raise albedo by as much as 0.4.

,. .. Figure ' Sb show's -the temperatu re· · chianges resulting from th~ art>~.do modification ·with respe~t ' f p1 ;the base·~c-ase simulations for 9 a.m. As ·show;h; j ri,

· r ··· " ·'f·;. f, . r ~ Figure Sc, at noon the peak ~impact has' moved slightly · inland, now centered over downtown Los Angeles with a cooling magnitude of_ 2~C. The-peak impact occurs in the early afternoon. Figure Sd illustrates that this im;~~ct excee9s 3°C at 3 p.m. We have conducted similar simulations under various irlitial coliditions, all of which indicate a peak summertime temperature impact ranging from 2 to 4°C., If w~--~o~e 6T = 3°C = 3K for 6a = 0.16, we can write Eq.(2) above and find encouraging agreement with Eq.(1) for White Sa11ds.

;.'r .. \ . )'." ;. ~ • ) r ~~· ·}, '. • ~: ' f i : ,· .. ~

19. ; Coftdusio1f' ' .. ,. -

- ( I Q•"' ' .- " ' ~ .. ~· ~:· j~' • ··~. \I (1,1 ;"';. • o ~•I l..) \,,.• t f

Using white surfaces to iricrease'l he"·albedo of ufban heat islaI}ds'~ay ~an easy way to conserve energy, save mol\ey i nd · reduce pollutibn. : Experim.ertts' have shown 20-40% direct saving~ by increasing the albedo of a single buildiQ.g, and computer siirnil-ation indicates that the' irtdirecfeffects of wiife..:s'C:ale albeCi~ 'chang~s will ne1a·f·1-y ~bre the direct 'savings.~(, .. , (. . . . . ' ; i '· . ' ~ .

l . :!~ ) ! <l","' ~ · ' · ' • i'J • • ~ · · .... }~ -:· =.". ,· .. '

Utility-~ponsored fnsM programs" to promote white suff_~_ces and trees ~ould earn about $100 million/year. Fof r'esidential buildfngs, asphalt shingle! and modified bitumen roofing should be targeted, and built-up roofing and modified bitumen are important in commercial applications. In most c~s,es, the albedo of these roofs may be changed at no· additional cost, at the time of required maintenance. L~beling programs are needed to identify . the best m,aterials for increasing ·aloedo,' and incen~ives are needed _to encourage ~he; i!lstalla~i,dn of light-r.oloted pavement. · ;y,.~ .. , ·

1 11 u -~ ' ... · ''3 :u ;· .. i ,,.· .,.,.,, •• .!.:;.... ..-

' ' ,,. ; l • . ._ I .. . : ~

'• 1. • :·. /ucr3/bealh_ahaB/(:IEE/1993.ICIEE .PAP/1~pcft>

!.

·.\Acknowledgment

.. . ....

:-10-

·. .J \ .., _ -• l

·• ' • • • \~ ~. (: :.i •. ~ •• . . - ~ r :;:~

,. This ~ '.w.ork. was supported by the California Jnstitute~ for Energy Effidenay ;(CIEE) through the U.S. Department of Energy, under contract DE-AC03-76SF00098. We

,, .. _,~e~.~owledge ~he contribution of 1many m~inbers of the Heat clslandi'> Group·at LBL, . ·. and .~specially thank Debbie. Giallembardo fo11.her ecjiting and production: assistance

and Lynn Price for her assistance with the figur~~.iiro •· ·'. •· , 1s · ,.; :- --

• ,•' <t:f'· 0 1..>.'J f. • _; ~1:

References

' ·~· ,,

.... !;: ';:...: .. , ...... . t ...

Akbari, H., Rosenfeld, A., and Taha, H. 1990. "Summer Heat Islands, Urban Trees, and White Surfaces," Proceedings of American Society of Heating, Refrigeration, and Air Conditioning Engineers, Atlanta, Georgia, (February); also Lawrence Berkeley Laboratory Report LBL- 28308.

Akbari, H., Rosenfeld, A., and Taha, H. 1989. "Recent Developments in Heat Island Studies, Technical and Policy," Proceedings of the Workshop on Urban Heat Islands, Berkeley CA, (February 23-24).

Bretz, S., Akbari, H., Rosenfeld, A., and Taha, H. 1992. "Implementation of White Surfaces: Materials and Utility Programs," (draft), Berkeley, CA: Lawrence Berkeley Laboratory Report LBL-32467.

Competitek/E-Source 1992. State of the Art: Space Cooling and Air Handling. Table l.SA estimates U.S. peak a/c (including air handling) as 210 GW, and 1990 sales of 420 BkWh (equivalent to 210 GW for 2000 hours). To estimate the hourly cost of 210 GW, we note that the marginal cost of peak power is 20-50¢/kWh. Thus PG&E quotes 33¢ (San Francisco Chronicle 5-12-93, Bus. Sect, p. Z-6). The 210 GW would then cost marginally about $70 M/hour, and our estimate of 10 GW (5% of 210 GW) from heat islands would cost several million dollars per hour. To convert from hourly to annual marginal cost, we assume that this very expensive electricity is used for about 1000 hours/year. Competitek has changed its name to E-Source, Boulder CO, 80302 (303) 440-8500.

Means Company. 1990. Means Building Construction Cost Data, 1991: 49th Annual Addition, Kingston, MA: Construction Consulting .and Publishing.

National Roofing Contractors Association. 1990. Miscellaneous sales records provided to LBL. NRCA, Rosemont, IL, (708) 299-9070.

Parker, D.S., Cummins, J.B., Sherwin, J.S., Stedman, T.C., and Mcllvaine, J.E.R. 1993. "Measured A/C Savings from Reflective Roof Coatings Applied to Florida Residences," Florida Solar Energy Center FSEC-CR-596-93, February 1993. FSEC, Cape Canaveral, FL 32920.

/user3/bel/h_ahag/CIEE/1993,CIEE.PAP/1uper6

--11-

Rosenfeld, A., H. Akbari, H. Taha and Bretz, S. 1992. "Implementation of Light­Colored Surfaces: Profits for Utilities and Labels for Paints," presented at ' the American Council for an Energy Efficient Economy 1992 Summer Study on Energy Efficiency-,fruBuildings.,.:Asilomar, CA. Washington, DC/Berkeley, CA: ACEEE.:

·: _ • • r ~ "-l >~ .: .~~ < .·· : :i :.: : : r . · · Lt ... : i ~ i

Sailor, Q.J;~·· a:nd K~ssler R. (1998):1~ _ .. Chapter 4.3: Three < Dimensional. Simu-fation Results, •i. in ,Analysis of Energy Efficiency and !Air Quality; .Interim Report, pp.'· 4.57-4.82, Lawrence Berkeley Laboratory Report: LBf.-.,.33051. . · · · ; ·'nn : ,- '.'"! · ·· ,·

Taha, H., D. Sailor, and Akbari, H. 1992. "High-Albedo Materials for Reducing Building Cooling Energy Use," Berkeley, CA: Lawrence Berkeley Laboratory Report LBL-31721. . : .. · · ·--t:r'

' A ~ •. ,i '· .

l ·; ... ,: • ~; , 1

1 )

. .

:··

~ _,

' ' ,, .;

-;-• ... ;'

. " . ' ' '1: ~ I • • • t.. • -

- I

," .. '

. . , . , . ~ .I ~· L ..

- 1'

~i

.J (It

)

'"'' ··

l ••

•l

I

' .! ~ r

~'

• ·e. ..

..

.. ·-

. ~ ' I . f . · .

~·

H 1•· " ­• !.>J

·'

f • .. ~ ~~ . ' ··.:.

,, -- , .. ,o ' 6 I ( ' . ..

.... • i.

··' ·· ·· ·

·,~ . . . ·~ .

,·. "'l" : .... ~-;: t i \

~·' d

1

.,

'""' I I,;~ :) ~,. ,_

: ~- -' r;i .'!. ' •

,. C I

• ':) • J

•. ;.t · ~

- 1.- ' - :...:. : -~1 r _ •

'· .. c;'": " · i · i.L [j, .. -., .. ' .... ,

. -~ · . .,.

. J .

·.r'-~' .~ 01 · •

, ·~ . . . r. ; t • •. . - . . ,.., '

.. ' •

.. :; ~. -:- · .... )' J ~ i-/

. ·, ~ , . "' .. 1,.1, : .

. ' -~ .'

' !

:.·I

I • ~ •

• . . .

•I

':. .: ,

/uaer3/bea/h_ 1h1g/ClEE/l 993,CIEE.PA P /•upc:i6

" ...

~'.' ·:· J~'- ... .... . •'

.t.; .~

"· t ·:

Ten-Year Running Average Yearly High Temperatures (°F)

105

104--::: -

103

102 -

101

100

98

97

1880 1900 1920 1940

Year

1960 1980

·­..

Figure I. Ten-Year Running Average Yearly High Temperatures in Los Angel~s, CA (1882-1984). With increasing irrigation and orchards, Los Angeles eooled 4 °F/yr until the -1 930s. Then, as asphah replaced trees, Los Angeles has wanned 7 °F (1°F/8 years). The ten-year running ave~e is cJc~1ated as thri·. average temperature.pf the previous 4 years, the current year, and the next 5 years. ·'

.. . tr' I f,

· ·~, . ·~ :;, I·

·;. ! " n; . .

•. I ·' ·.-,

. r . . ·

. "

• ';. l .

\' l

,, .. .. !''

'I

.. ': "l . .I : ....

• • f; . ) .

' .

'!

-..

Peak Power (left scale)

9~----

'. · .. :; 'J :; \ . . .. '

·~

f <'"" , -. .

65 ·- ·

j _ ....

70 75

~ I

Smog (right scale)

' • :i''-' I

-· ) . f/ .. 11

.. ,: ... • i..'i.: • ~ I ' •. • ~ . ·L

80 85 ) -

Daily Average Temp °F

.. !

·' ~\_' \

. . .... • ...

',}

1',

"'~-

"".

t •o # .. ,.

r, I . 'l -9K . ' '··' 100

Figure 2. Ozone Level and Peak Power in Los Angeles, CA. Below 70°F there are almost never smog "episodes" in Los Angeles, i.e. the ozone level stays below the National Ambient Air Quality Standard (NAAQS) of 12 pphrn. Starting at about 73°F, smog episodes begin and exceed 50% by 90°F. In Los Angeles, summer temperatures average 78°F. If the 7°F heat island were eliminated, the average temperature would drop to 71°F and smog incidents would greatly decrease. The probability of smog is derived from a scatter plot of ozone concentration vs. daily maximum T (Figure 6 of Akbari, H., Rosenfeld, A., and Taha, H. 1990. "Summer Heat Islands, Urban Trees, and White Surfaces," Proceedings of American Society of Heating, Refrigeration, and Air Conditioning Engineers, Atlanta, GA (February); also LBL-28308).

0

.. . -..,

~ ~·

·:.1 ('' •'

'.t; , J .... .... , 1 ... •

.. ·s· ~:l

_.,,.

•' ,

14

· <, r Surface Temperature 0c r · Surface Tpmperature °F

• 4

/ . 90 ....--------------b-la_c_k-si-ng-le-pl-y--...,..,b-la_c....,k·. ;190 ./ ;,, C9ating

~ ·'., , . ,1_, •• ;' ' • ·'

v // 80 ~ . ',. ~ .... ;---::./ .... · ,...

60

501 ,, ,,. . ,

./," .. . . ·:.>.!' ,, , ...... ~ ..

40

30

/ .180 AVERAGE CITY.! , "white" asphalt shingl~ ~URAL A.REAS

aluminum coating·.o • '~ ? terra cotta tile 170 " . . ~alvarnzed st~t t',~~r & qravel " . ' .. , -. J, . ·~~ , I . ' ' ., '· ' . • . . •'g~ ... ~ I ~YPOTHETI~~~: GREEN CITY f wea,h:;{ed concrete

" ·. /I •, ' -· \Ii I :

t:t'(;l~OTHETICAL WHITE CITY... 1

I. ·~ 0 / I I 10 · ,, • .. . . I ~ ~ aluminum coat~~ ·•very light gray

~ · /:light beige

e I ', :aluminum foil

I I

/

/ I

. ,/j white paint, • 1 / anodi~d aluminum

I / I

- - -ambient air

I I I ..... I I

-'­I I

.... "! ' / ,

-·"'-.. .. .. ,

150 ·,

'. f4o

130

120

110

_TOO

90

0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 Solar Absorptivity (1-albedo)

: ~ T ·. ~"": t· ~ i • • . . '

·~ · · • I' ...

'Figure 3. iSolar Absorptivity .. vs. Sur.f'ace 'Terriperatqre:of Horfi ontal Surfac·~~; !Jiaint~~, ro·ofing : m~terials, i'oad'_Vays; .a~d.'1:il1~s, adjust~~}O noon on '~ deaf ,"~i.~dle1~S sumirter ~,~ .iJ:1 ;\us~J~ /rX. ,, : · _ ~ere are large t.~rt'~e~ture ~~re31?~ of'ab~~\ ~~~f~ ootr.~9 w~~~ - ~d: blac~ suf!~qe.s.~~ .9, 4Q~f; 1

be~;en con~rete· .an~ ~sphalt. ~phhlt .~ur(~~a ,~~ 19rus~~~ oyst~r ~~~µs: ~~ . .s~q ~. P!Qbably. 6Q°F cop~er ,'.than '!!'1e ~di~~a1 blac~ Yt.~'jon. There ~s ~~o a l.ar$~ itemeerature. spr;ead.betweeq~ alµ~um (or'Wh1te) arid galvanized steel. Hoth metals run hotter than_ paip~. because they radiate heat j>Qorly (have a low "emissivity"); in addition, galvanized steel has a high absorptivity.

As surface temperatures change, so does air temperature, but much less sensitively. Nevertheless the average city's surface temperature is hotter than its rural surroundings, and thus the air is warmed. The figure shows a hypothetical light-roofed "green city" with surface temperatures 20°F-25°F cooler than either an average city or surrounding rural areas because of the combination of white roofs, light streets and parking Jots, and urban vegetation. In the hypothetical "white city," there is less existing urban vegetation (such as in Phoenix whe.re scarce water means fewer lawns), and the surface temperature is funher reduced.

• We know that the absorptivity of aluminum coating ranges from 0.2 to 0.5. In one test, aluminum coating reached 172°F.

Source: Taha, Sailor, and Akbari, 1992 (LBL-31721 ), Fig. 6. and Table 11.

.. - 500 -

- .. , CD ···-' C °' 300 :::J

>--u !: 200 0 CD

~ .··;(_ gi 100

0 ("• 0 ()

IJ

·, ,•

O-f'-~~-----,----------ro--------:......;~------....;...~~~~~::t.-0 >~ =· () as as c;:c -. ·, ,_; ..

CD > as ... a

. "ed . -

-. • •I )

0 •. 2 CD

-"C CD ...

0.4 "<f 6 Q) -CD ... u c Ro.of albedo 0 u ··

i: F.

.....,...--., ,.

. j\ .. -.. ' •

0.8 ~ . 1 t.)' - > CD c ----as CG .c a. - :r: 0 CD ----.c ,-~ ~~ ..

' ... \,;

70

60

50

40

30

20

10

0

Figure 4. Effect of Roof Albedo and Tree Shading on Cooling Energy Use for a 1960s House in Sacramento. The lower solid line labeled "Many trees" is a fit to a real, well-shaded h0t~~.e., .. ~u~ usiQ..g .~ir1m11 ~~~ther data . (!1~ Jpc~J c,ooling.fJ?Y. t_he tree~). Tbe·· cross-hatched. ~ea repr~.s~pts , a, rang~ o~~"vmgs frow Jfe~.s .. 11 A,t the ~~p W:<; ,e~.umate . the ~ useJby the--§ame h0use "".ith j.o'st'·'-j treees' and no <;~oling 'of th~ mi~rocljplate .. At _the bqttP.pi .we e~timate the niiciocl~mat~·effes;t of dense 'fr~.~~ · (B ~it~ir g ·f qn~il:ions, ~uiq· .~~!11ulatioo . a~s.~mptjons: ,Floor .w-ea 17~01 fiZ:Lone .~tory; .~.11 ·r.c:J~f. ~7 .~,al!s.~ · ~O _qoo( qoubly' ~!~i.ng; theh;ri.ostat S<i~poi~L~0°F; cooling·ISEER 12.6, ayetage · res1dennal'~lecmc ra~e for California $O.q5/kWh . • i$ource: ~EC Energy Watch, V,ol. 14; No. 5, June 'l992). : .: 1

' ' • • ". • • f' \ q

61 ...

... ,• I '•l.

. , ; :' . ..

··' ,·.1. . ....

·'" ..... \ .. ~.

'· .

. .f • .

-'-m Cl> ~ .. --0

0 0 T"" ...... ~ --en 0 CJ

C> c: c: 0 --'O c: 0 CJ I ... <

30

-....... . ..r:::.

~20 .._. '

Q) ..... ·-en . .... Q

. 10 (..)

• <(

~ -·ca Cl

. . ...... . ... - ... - - ••• •• • • • •• 'II .... .. .. . . . . . ..... .. . .

.~ -~ 10 C.;

0 0 0 0 .·· . .. . ·

,_ . .. .. · 0 b0b ......

,,...lb! ..... •• • : .:·.:.

"'

·'3''.:···· . . .. ... . .. : . ' . 0 .·· . , r.:..~ •• ·· . : '7 .. ·· 1: .. ) o •• • •Y

' .. ..... .. · ··-~ . ·. !::. •..• ···•• .

~ 0 ..... .... .~····

.. :>:·: ... .... .. .... : .......... .. ...... ..... : .. . '

"·

I ' . .

,_.., u,,.

30 Oaily1LA)c~. for site .2 · (kWh I day)

Figure 5. Impact of 8 Large and 8 Small Shade Trees on 2 Houses in Sacramento. Squares ~d their sou~ regression line repr~senc 19 July base case days with no trees. J:?iamonds show ·~ tfoxt 2~ August days widi' Sife .. 2 shaded. These· data and Ol'eir ·regression· m0V.e left ; abd~c J3%t · Triangles .. show 39 Sept~rtit>er/October days with :the shading moved t0 Site· 1 ~ ;These data move down, agath· by about 3S%i ;~hading saves about 4 kWh/day:. . · · · · · " ··

•' .. 14 :· ·· · ··· ·········· · ·· · ·· :···· ·:· · · ··- ·· ·· ··· · ·· ···· ···· · ··· ·· ·· · -· -· ·· ·· ·

'. I ~ .. 1 • • • • • •\." ~

: . ' : : ~ :·"""-.-. 12 ~- ... . .. ..... .... .... . . . ~ ... .. · ... . . . ... . '. .. -.. .. . ~- .. .... .. ... .. . ... -l · : ..c: 3: ~10

>. O> '-Q) 8 c: w C> .5 6 0 0

(.)

~4 ·ca c

2

.. : . : 0 : : \ .r,y, : : : : .. ..... ," .. .... ... ...... .. : .... .... · ... .... -... .... : ... ... .... ·o· .. . /. .. -< .. . . . . . ~ : · ... ~ /' : : o I .. ~. . . ·. . : ~ 0 0 . .. .. . . ... . - . . . ........... .. . .... ~ .. ..... . . . . . .. .. . . . .... . . . . . . ..... . . . . ... .. . . . : : -~.J . . .;..

. : ; · ,;. :

.. ...... d .--. .... ........ .... .... .. 1 ( . &I o .... .. .. . .. .. .. . . . . . . . . ...... . ~ ... . .. .. .. . ' . . . ' .

,•· .......

23 25 27 '. . .. . -~ ~ .. , 0 ., ... ,

. j ' Daily Aventg,e' Temperature (~),

ll · f,r, , · ·?, "'. = • ,~ ' ·.. !rf·~ ·: ~ , . . .·i·J·- ~ ·: ',·q: f ~ .;r:.f

Figure 6. Daily Cpoling-E11e~gy Use fl~ . Site 2 Versus ~~il·Y· A v.erage Outd09r Te111per~~~·~· ·1'he squares ;reprt~nt 1991; pre-modiflcatipn conditions, c.wilh :~: .roof albedo of 0,.1 8. The circles, many of them at 0.0 kWh, re!'resent the post-modification~riod, with an alb,¢do of 0.7~1 ..

"C c

"' E (lJ

0

-"' U · -t.. ..., u (lJ

4000

3000

2000

u; 1000 u <

0

,. I I

I I I

.o . .

,, , .. I 1

:: r· _, .. ··'

. ..... ..

\ .. ,,,,.

,.

( '• } l'. r1 f ·v .' '-1 l. , I !_ j l ,

1~ /': Ro/atea. July 31st

'\ '·~ I' t I I ' ; 1 , Root a1,space te11p

\ \ f .. ' .. ' • • ' \... .f .() ' <f \ ..., ~ - I '~ ~ t 1

, • I \.

........ ' '·~' "' ... :...,' ·'; ... 1:·-;:.1 .. , .. _,' •-,. •

72 96 120 1~4 - 1Ei!3 _ 192 Hour of Week

SITE #2~ WEEK OF ROOF COATING

•' \

120

110

100 ~ OJ

90 c... => .....

BO IC L OJ

70 a. E OJ I-

~--IC a. If) c... ..... < -0 0 a:

'- I.

' . '·'

.. )

Figure 7. Roof Air Space Temperature and IS-minute Air Conditioning Consumption at Site 2 from July 28th to August 3rd, 1992. The roof was treated with a reflective roof coating

.. .. -· on ~µly 3lg ..• . B~h ,roof temperatures and .cooling energy consumption were .substantially :· reduced .. ·~. cpnaii"pning el~ctricity use· was. decreased by 43%. over,·perioos 1with similar . - weather con~tjQn~ (Parl,cer et al.' 1993). . ' : . . ''"~ : . : . ) ., '.

!

j.) ....

c:

Q) L..

:J

cu L.. Q)

a. E Q) -

110 Source: Taha & Rosenfeld, 1992: Work in progress, LBL

C - Station (albedo= 0.25) r--1 - - ---

/ ---.J---- ------,-·-- _ .. -··· ---··· --- .. ··-·-/ '

105 -1---I ' ,' ,,---------- ---- -- -,, ' ,,

I ' \ ' I I ' '

,' / Northrup strip \ ', I / \ ',

I , \ ' I / \ ',

/ ,' '-, ', I ,' ' - · ·'- -

___ , ----~ . .. . \

/ I ' , ' ---100 -· . --/ -/ · . . . . .. . .. -- ------~""""' ---.,,,-----,: .. ~-- ,-, '""'\', I I ' • .,~.. \ \

I ,' ", ,.,,. ' ' \ , \ ' J I • '! \ .,,.,,,,. \ \

I ,' \ .,,.... ' \ I I •' \ '

I I \ \

I ,' ' \ I I • • \ • \

/ / ,_.. Ratscat .(.albedo= 0.60) '- \ I I " \

I ,' ' \ I I '" ',,! ' \ '

I - ' \\ \

95 / / '· \ \ I I \ \

I ' ·'"' '.- ·: \ \

~ \

90

I I

5

c .I.,

I .

6 7

"'Month, 1991

J!

9

' ' ' '

'

' ' ' \ \

10

E: r, .~u~)~.: ·; ~"'} :-: : 7• '..: I 4"' '

~ I:U.gure ·Ad vi· 1991 :Montkly Maximum Temperatures, White~>S'an'\ls · Mis&ile Range':· N~w :-i • ;1 Mexico . .-Ratscac :weather:Sta:tion is ar the edge of the white gypsurrViirea: NorHiiup and C-Station

mea~ure th~ surrounding temp~rarur~s. hut North mp is <:: loser t0··· he fringe· bf the white, and should be cooler than C-Station.

3870 3800

(b)

3821 3775 {7

E -C$,3773 '

E e375o * Riwfside

>- >-

3724 3725

3675 3700 280 360 440 520 600 350 388 425 463 500

X (km) x (km)

3800 3800

3775 3775

e E c31so e375o

>- >

3725 3725

3 7 00 +-.....;:::.--...:::;.._,...___,.__,__,,_.--..--~ ......... ...,..:--..---.----.----..--..-..--.,...........1 3700-+-~--~__,.;~~--..-.-.--~---....,...;..~~-.---.--.---..~~

350 388 425 463 500 350 388 425 . 463 500

x (km) X(km)

Figure 8: Albedo modi fication results: ( a) Regions within the modeling domain which have been identified fer simulated a lbedo augm entation. Grey is unmodified. hashed is a modification of less than O .. to. an4 white is• modification in excess of 0.10. The average albedo increase over the 394 modified cells is 0.18; (h) 1.emeerature d ifference between high-albed.o case and base case simulation at 9 a.m. (c) same as b except at noea: \--4) same as b except a t 3 p.m. Contour increment in b-d is 0.5 deg. C.

t-J 0