08650-3-T THE U N I V E R S I TY OF MICHIGAN COLLEGE OF ENGINEERING Department of Meteorology and Oceanography OBSERVED TEMPERATURE PROFILES NEAR THE LAKE MICHIGAN SHORELINE Dirk Herkhof Aksel Wiin-Nielsen, Project Director ORA Project 08650 Supported byDEPARTMENT OF HEALTH, EDUCATION AND WELFARE PUBLIC HEALTH SERVICE DIVISION OF AIR POLLUTION GRANT NO. AP 00380-01,02,03 WASHINGTON, D C. administered through: OFFICE OF RESEARCH ADMINISTRATION ANN ARBOR January 1969

ACKNOWLEDGEMENTS The author is greatly indebted to Mr. Anders Daniels for suggesting the details of the circuitry used for the instrumentation, and for his many helpful suggestions during the execution of this study; and to Prof. G. C. Gill for his assistance in selecting the instruments. The writer also gratefully acknowledges the work done by Mrs. Ruthie Tolbert in typing the manuscript. Particular thanks is given to Dr. Alan Cole and Mr. Lars Olsson for reviewing the manuscript and making many useful suggestions. ii

TABLE OF CONTENTS Page LIST OF FIGURES.o.o o.. o. o o o.... o iv ABSTRACT......... o o.................. vii 1. INTRODUCTION....................... 1 2. LOCATION OF STUDY AND MEASUREMENT METHOD.............................. 3 2.1 Description of Location....... 3 2.2 Instrumentation and Setup.... 5 2.3 Operational Procedure. o...... 6 2.4 The Time Constant... o o...... 8 2.5 Calibration................... 9 3. THE DATA AND THE RESULTS.......... 9 3.1 Results of July 2, 1967; Prevailing Westerly Flow...... 12 3.2 Results of July 6, 1967; An Ideal Lake Breeze Day. o... 16 3.3 Results of July 8, 1967; Weak Southwesterly Gradient Flow... 27 3.4 Results of July 16, 1967; Weak Southwesterly Gradient Flow o.. 31 4. SUMMARY AND CONCLUSIONS. o o....... o 34 REFERENCES o o................ o.... 36 iii

LIST OF FIGURES Figure Page 1 Location of observation stations on the eastern shore of Lake Michigan.eo.. o............... 4 2 Circuit used in measuring temperature..... o o o o 0. o 7 3 Ventilation pipe with radiation shield containing thermister probe mounted on automobile O O o. 0....0 7 4 Calibration plot of the recording device. oo o.................*. 10 5 Full scale deflection of the recorder as a function of mid-point temperature............... oo... 11 6 Mesoscale surface winds and temperatures around the southern basin of Lake Michigan at 1300 EDT, July 2, 1967.. o o o o o o 13 7 Westerly and southerly components of the wind at the M-45 stations for 1000 - 1400 EDT, July 2, 1967. 14 8 Temperature distribution observed 1230 - 1355 EDT, July 2, 1967 and the winds at the M-45 stations at 1300 EDT. o o.. o. a. a 0 0 0 0. o. e o o. o o. o. 15 9 Mesoscale surface winds and temperatures around the southern basin of Lake Michigan at 1600 EDT, July 6, 1967................... 17 iv

LIST OF FIGURES (continued) Figure Page 10 Westerly and southerly components of the wind at the M-45 stations for 0800 - 2400 EDT, July 6, 1967.......... 18 11 Temperature trace along M-45 for 0945 - 1005 EDT, July 6, 1967... 19 12 Temperature trace for 1220 - 1240 EDT, July 6, 1967.......... 19 13 Temperature trace for 1415 - 1435 EDT, July 6, 1967.......... 21 14 Temperature trace for 1435 - 1502 EDT, July 6, 1967.......... 21 15 Temperature trace for 1526 - 1545 EDT, July 6, 1967 o.o......o 22 16 Temperature trace for 1545 - 1610 EDT, July 6, 1967.o o.o... 22 17 Temperature trace for 1617 1640 EDT, July 6, 1967.. 0.... 23 18 Temperature trace for 1640 - 1701 EDT, July 6, 1967.......... 23 19 Temperature trace for 17101730 EDT, July 6, 1967 o o........ 24 20 Temperature trace for 1730 - 1750 EDT, July 6, 1967. *. o.o.. 24 v

LIST OF FIGURES (continued) Figure Page 21 Temperature trace for 18381855 EDT, July 6, 1967.......... 25 22 Temperature trace for 1957 - 2015 EDT, July 6, 1967......... 25 23 Temperature distribution on the lake-breeze day of July 6, 1967..................... 26 24 Mesoscale surface winds and temperatures around the southern basin of Lake Michigan at 1200 EDT, July 8, 1967................. 28 25 Westerly and southerly components of the wind at the M-45 stations for 1000-2000 EDT, July 8, 1967.. 29 26 Temperatures recorded along M-45, for 1130 - 1203 EDT and 1810 - 1845 EDT, July 8, 1967.......... 30 27 Mesoscale surface winds and temperatures around the southern basin of Lake Michigan at 1200 EDT, July 16, 1967.................. 32 28 Westerly and southerly components of the wind at the M-45 stations for 0800 - 1800 EDT, July 16, 1967 32 29 Temperatures recorded along M-45, for 1120 - 1200 EDT, and 1740 - 1755 EDT, July 16, 1967......... 33 vi

ABSTRACT A temperature measurement program using a continuous recording device on an automobile was carried out near the shore of Lake Michigan south of Grand Haven, Michigan, on several favorable days in the summer of 1967. This study provided a detailed illustration of the temperature distribution along a line perpendicular to the shore on mostly sunny days and under different prevailing wind situations. Of the four days during which good data were obtained, one had a prevailing moderate westerly wind, another a prevailing easterly wind with a lake breeze gradually pushing inland from the westo On the two other days light southerly to southwesterly winds prevailed. A description of the instruments used and a brief account of the operational procedures are given. The temperatures have been plotted as a function of distance inland at various times of the day together with the winds recorded at different distances from the shore. vii

1. INTRODUCTION The modification process of the air in the lower layer of the atmosphere when it crosses a sea-land interface has received increased attention recently. The effects of the land-water transition have many important implications for the atmosphere near the shoreline of a large body of watero Since the process is very complex, a theoretical study is not a satisfactory approach to the problem. It is therefore necessary to carry out an extensive field program to measure the rate of change in the physical properties of the air as it undergoes modification. An ideal site for such a study is near the shore of a large lake in the interior of a continent in middle latitudes in spring and early summer, when the lake is quite cold with respect to the surrounding land. The generation of a lake breeze on a calm, clear day offers an especially good opportunity to observe the modification in temperature of the air from over the lake as it moves inland. For several years, the Department of Meteorology and Oceanography at the University of Michigan has maintained an observational program on the eastern shore of Lake Michigan to study mesoscale winds in general and lake breezes in particular. Instruments at stations at the shoreline and at different distances inland have continuously recorded wind speed and direction as well as temperature and humidity. Pilot balloons have been released to study. winds aloft. Results of these studies have been -1

-2presented by Moroz (1965) and by Olsson Cole, and Hewson (1968). Lyons and Wilson (1968) have observed that during onshore flow in non-lake breeze situations, the eastern shore of Lake Michigan is cloudless, while cumulus clouds form about 25 km inland as the air gradually heats up on passing over land. Air temperature modification over a lakeshore has also been discussed by Bierly (1968). In the modification process, important changes take place in the stability and wind of the air in the lowest layer, thus profoundly effecting turbulence and diffusion near a shore as shown by Hewson and Olsson (1967). Thus the lake influence has important effects on local winds, cloudiness and convective precipitation, air pollution, and local climatology. Thus it seemed appropriate to carry out a study in the summer of 1967, to provide a more detailed analysis of the air temperature near the lakeshore by using a continuous recording device on an automobile. This study had several objectives. First, to find a way of more clearly following the progress of the lake breeze front, i.e. the line of penetration of inward surging lake air, by noticing where a discontinuity in the temperature gradient is located. Second, to determine a typical temperature distribution near the lakeshore during different situations of prevailing wind caused by the heating of cool lake air when

-3passing over land both during lake breeze and nonlake breeze situations. The second objective would indicate how fast the air is modified and how far the lake effect on air temperature reaches inland. Automobile traverses have been made use of in the past by Sundberg (1950), Duckworth and Sandberg (1954), Summers (1965), and Daniels (1965) but only for measuring temperatures in cities Few if any, traverses, to the author's knowledge, have been made anywhere near a large lake or ocean. 2o LOCATION OF STUDY AND MEASUREMENT METHOD 2ol Description of Location The study was carried out on Michigan highway M-45 which runs east-west from Grand Rapids to the lake. The shore at this point is only slightly curved and runs north-southo Along the highway are 4 meteorological stations set up by the Department of Meteorology and Oceanography at the University of Michigan. They are located at the shoreline, 8 km, 16 km, and 24 km inland. Each is equipped with wind sensors on a tower, which continuously record wind speed and direction, in addition to a hygrothermograph which records temperature and relative humidity. Instruments located on WJBL-TV tower, 17 km south of M-45 and 10 km inland, record winds at 3 different levels and temperature lapse rateo The location together with the stations is shown in Figure lo

MUSKEGON O MKG L KE GRA\END _MILWAUKEE HAVEN GRAND l AVEN CHIC LAKE tRAND MICHIGAN LAKE RAPIDS SHORE 8KM 16KM 24KM M —--- __S~^ 45 GRR 0 10 KM - OWJBL -' FIGURE 1. Location of observation stations on the eastern shore of Lake Michigan. Insert shows the location of the line of observations relative to Lake Michigan.

-5Sand dunes, with heights of up to 50 m above lake level, parallel the beach and extend 1 to 2 km inland. Beyond 3 km inland, the land is flat with relatively few trees and is uniformly developed for agriculture. The slope of the land from the shore to Grand Rapids is negligibleo 2.2 Instrumentation and Setup The temperature sensor used was a Yellow Springs number 408 thermister probe with a time constant of 0.8 seconds. The Wheatstone bridge circuit is sketched in Figure 2. By setting MR, one determines the temperature of the midpoint of the scale, which allows for days with differing mean temperatures. The potentiometer, MS, regulates the voltage across the bridge which determines the sensitivity - the amount the pen will swing for a certain increment of temperature. The circuit was powered by a lo5 volt batteryo The circuit was built into a small box with two dials to regulate the sensitivity and the midpoint temperature. The box had three outlets: one for the recorder, one for the probe, and one for a switch by which the circuit could be opened or closed. An Esterline Angus Model AW recorder was powered by the 12 volt car battery using a 115 volt, 60 Hertz AC inverter. Artificial ventilation was provided by an aspirator in the lower end of a 4 foot vertical pipe with the thermister mounted at the top; the pipe also served as a radiation shieldo The aspirator was tied to the inside of the right

-6front door, and the pipe was pointed through the open window straight up so that the thermister was located one or two feet above the top of the car, Figure 3. The probe was pointed into the direction of motion of the car. This was found to be the most favorable position for the speed of the car didn't seem to effect the temperature recorded, and no vacuum effect was present. 2.3 Operational Procedure In all runs, all operations were carried out by the driver. The rides always started at the shoreline station where the temperature was read from a hand carried Yellows Spring meter using a probe by Yellow Springs, model 403 mounted into the pipe next to the regular probe. The local time and the indicated temperature were marked on the chart. The range was adjusted so that the pen would stay on scale during the entire run. At 16 specific locations - each about 1.5 km apart - markings were made on the chart by breaking the circuit with the switch. The run usually went 20 km inland, where another temperature reading was taken and recorded on the chart, which estabilished a reference for abstracting the data. Since the runs took only about 20 minutes one way, diurnal temperature variations could be neglected.

-7-, RB i a 2 o r a a a — 1.5 V pM wia~e c tann ms RECORDER. MR FIGURE 2. Circuit arrangement for measuring temperature. Ra and Rb are constant resistances of 1000 ohm each, R is the Yellow Springs No. 408 temperature probe, PR is a 270 ohm resistor acting as a voltage divider. Mr is a 50,000 ohm potentiometer which is used to adjust the midpoint temperature. Ms is a 100 ohm potentiometer used to adjust the sensitivity. FIGURE 3. Ventilation pipe with radiation shield containing thermister probe mounted on automobile.

2.4 The time constant The time constant of the thermister is supposed to be 0.8 seconds. However, because the recorder also has a time response, the system response is slower. The real time constant was determined by putting the probe in a small wind tunnel and exposing it to abrupt changes in temperature, and recording the response of the system. The time constant is defined as the time it takes for the system to respond to 63.2% of the total instantaneous change. After several tests, the time constant was estimated at 5.1 seconds. The time response formula is T1 - T = AT e-t/ T where T1 is final temperature, T is instantaneous temperature at time t, AT is the total change in temperature, and T is the time constant. It follows then that t = -2.3 T log T1 - T. AT Then for 90% of the change to take place T - T = 0.1 AT and t = 11.7 seconds. If the car travels at 40 mph during the time that 90% of the total change is recorded, the car travels 210 m; at 60 mph it travels 314 m. Temperature gradients of 30F per km have been observed at times near the shoreline, where the speed usually was 30 to 40 mph. Therefore, a lag of 210 m wouldn't effect the accuracy significantly. Inland, the temperature gradients were much weaker. The response was still so fast that the trace showed a lot of -8

-9small scale fluctuations in temperature, often of the order of 1~F in less than 1 km, thought to be due to turbulent eddies and local effects 2.5 Calibration The system was calibrated using a sensitivity of 5 and a range of 90 on a 100 unit scale. Calibration was carried out by dipping the thermister in a water bath using different temperatures in a range from 60 to 80~F, the temperature range which usually existed during these studies. The response seemed surprisingly linear in this case as shown in Figure 4. Thus a linear scale was constructed and its validity could always be checked against the observed temperatures taken during each run. In Figure 5 is given the full temperature range for different midpoint temperatures (the temperature at the center of the scale)o 3o THE DATA AND THE RESULTS Good data were obtained on the following days between the times (EDT) listed: July 2 1230 - 1355 July 6 0945 - 1055, 1220 - 1305, 1400 - 2118 July 8 1340 - 1203, 1730 - 1845 July 16 1120 - 1200, 1740 - 1820 All runs were made during mostly clear and sunny weather, under a variety of prevailing wind conditions Westerly winds prevailed on July 2o

80 SENSITIVITY = 5 RANGE = 90 x UxX IL Lx 0 Xo z x 10 20 30 40 50 60 70 80 RECORDER OUTPUT IN MICRO AMPS FIGURE 4. Calibration plot using a sensitivity of 5 and a range of 90 on a 100 unit scale. Midpoint is 71F. Full scale temperature change is 4.160 70 80 RECORDER OUTPUT IN MICRO AMPS FIGURE 4. Calibration plot using a sensitivity of 5 and a range of 90 on a 100 unit scale. Midpoint is 71.4~F. Full scale temperature change is 14.1~F.

-1124 22 L.\ 0 z 0 z 20 C) L \ X 16 \ ^ 0\ 12 0 I I I 40 50 60 70 80 90 MIDPOINT TEMPERATURE IN ~F FIGURE 5. Full scale deflection of the recorder as a function of midpoint temperature.

-12Winds were easterly on July 6, while an ideal lake breeze was generated near the shore. On July 8 and July 16, southwesterly winds prevailed. Wind and temperature data were available from the M-45 stations and the WJBL - TV tower. 3.1 Results of July 2, 1967; Moderate Westerly Gradient Flow The morning of July 2 was clear and sunny. A cold front had passed the day before and a northwesterly to westerly flow averaged about 4 m sec at the shoreline in the morning. Toward noon, however, the wind started to back toward the W and WSW/ Figure 6. Figure 7 shows the westerly and southerly components of the wind. At the stations inland, the wind increased during the course of the day, but at the shore the wind decreased. The sky was clear, but cumulus clouds were forming 20 km inland. Temperatures were taken along M-45 going 25 km inland, and along Port Sheldon Road 10 km to the south. The ride lasted for 1.5 hr, and since temperatures were rising due to diurnal heating, the data were normalized to 1300 EDT assuming everywhere the rate of increase was similar to those observed on the temperature records at the stations. The route followed and the observed temperature field is plotted in Figure 8. Along M-45, the temperature increased gradually at the -l rate of 1~F km from the shoreline inland to 10 km. Beyond that point, the temperature gradient was much smaller and approached the representative inland value. At Port Sheldon Road, isotherms

-13f(~~ \GffMKG:^-O I4GRR104 LAKE MICHIGAN WJBL BE " 74 119 5 94 =4 CGX/ *2 WDW </ 0DW T? KIT 5a 4* 1300 EDT, JULY 2, 1967 FIGURE 6. Mesoscale surface winds and temperatures around the southern basin of Lake Michigan at 1300 EDT, July 2, 1967. were packed closer together with a gradient of 2~F km1 observed between 4 and 6 km inland. Thus, the horizontal temperature gradient was not homogeneous along the shoreo This might have been due to local effects caused by differences in topography and surface covero Port Sheldon Road goes through gently rolling farmland with few trees. At 4 to 6 km from the shore there is an area of dense woods, and near the shore there are sand dunes covered with many trees, which might have been a factor causing the higher temperature gradient observed thereo At 1400 EDT, a long squall-line like cloud formation was seen in the NW, far out over the lake. At the shore the wind backed to SW and decreased, obviously due to the approaching disturbance over the lake. The cumulus clouds over land disappearedo

-14SHORELINE o. / -2 - -.I I....I. I 10 II 12 13 14 8 KM 2 o / 2 I 11.1. I I I 10 II 12 13 14 o 16 KM W 4 3 0 8' 2 I,I I I I..I 10 II 12 13 14 6 24 KM 4 2 o.I I I I I 10 II 12 13 14 TIME, EDT JULY 2, 1967 FIGURE 7. Westerly (solid line) and southerly (dashed line) components of the wind at the shore, 8 km, 16 km, and 24 km inland along M-45 for 1000 - 1400 EDT, July 2, 1967..

-151300 EDT, JULY 2, 1967 2 3 KM 1 65 66 67 68 69 70 71 ~ -- I I I I LAKE- _ -X- - ---- -X- -X- -X- -X- - -- -XSHORE M-45 8KM 16 KM 24 64 STATION SSTATION 66 \ PORT SHELDON RD. X- -x_ -x —x - x -X - X- -X- -x —x —x — x < | 65 6768 69 70 71 o- o NW y-0 —I t'^ —'WNW o_.- -___ c" WNW 2 — -o -— _o__________o~__: w __.~ -I~ bi~ -zsw - o 4 - 3o o o WSW 04 zn Z 2 0D I I I I I I 14 I I6 18 20 22 0 0 2 4 6 8 10 12 14 16 18 20 22 24, DISTANCE FROM SHORELINE IN KM FIGURE 8. Temperature distribution observed, 1230 1355 EDT, July 2, 1967. Wind direction and westerly (u) component of the wind observed at 1300 EDT. Temperatures are in ~F and normalized to 1300 EDT.

-16No more observations were carried out in the afternoon as a line of thunderstorms swept through around 1630 EDT. This was an excellent case of modification of lake air as it moved inland. The temperature increased gradually from 64~F at the shoreline to 69~F at 8 km inland, and from there increased more slowly to 71~F at 25 km inland. The U.S. Weather Bureau station at Grand Rapids, 53 km inland, reported 73~F at the time which suggested that the effect could have reached beyond 25 km. If the westerly component of the wind was 3 m sec-1 and a temperature gradient of 1~F km existed, the rate of temperature advection must have been about 10 to 11~F hr-1. Since the temperature at any particular point was rising 1 to 2~F hr1, the rate of modification of the lake air as it traveled over land must have been somewhat greater than 10 - 11OF hr1. 3.2 Results of July 6, 1967: An Ideal Lake Breege Day. A high pressure area was located over the Great Lakes with a low to the south causing a light easterly flow of about 3 m sec-1. As sunny weather prevailed, a lake breeze began to blow at the shore in the early afternoon working itself gradually eastward. The local situation is shown in Figure 9. Observations were made along M-45 all afternoon and early evening. Wind data plotted in Figure 10 indicate that the lake breeze started

-176e( MKGE MKE 52 \ M-45 GRR 7 o,,,,~ 243 LAKE MICHIGAN WJBL vtAZO ( / ^50 760RD 232 58- CGX MDW \ SBN 230 JOT 77 1600 EDT, JULY 6, 1967 FIGURE 9. Mesoscale surface winds and temperatures around the southern basin of Lake Michigan at 1600 EDT, July 6, 1967. at the shore at 1310 EDT, reached the 8 km station at 1730 EDT, and the WJBL station, 10 km inland, at 2020 EDT. The USWB station at Muskegon Airport (MKG), reported a wind shift by 1600 EDTo A temperature difference of 10~F between the air at the shore and the air far inland existed during the lake breezeo The penetration of the lake breeze front was quite slow because of the easterly gradient windso In Figure 11 through 22 are given some of the original chart traces from the runs made that day. (Notice that the horizontal scale is not linear due to the uneven motion of the car)o Figure 11 and 12 illustrate the morning and early afternoon case before the lake breeze startedo As easterly winds occurred everywhere, no lake influence on the temperature field was seen anywhereo The next run,

-18SHORELINE I /~ \/I,. \ -2 t 2 6 -4 - /LLAKE BREEZE I I I I I I I! I I 08 10 12 14 16 18 20 22 24 4 8 KM 2 /'& —_\A 0 __ -2 LAKE BREEZE \B / -4'., 08 10 12 14 15 18 20 22 24 4 - WJBL STATION, IOKM 2 — \ / \ F 10._ _We s/ ( —oi /a 3 -2 _ - A <\\ 4-4 AKE BREEZE 28 10 12 14 16 18 20 2 20 4 4 16 KM O -2 -4\.I_1,. I I I I [ I I I I, I I 08 10 12 14 16 18 20 22 24 TIME, EDT JULY 6, 1967 FIGURE 10. Westerly (solid line) and southerly (dashed line) components of the wind at the shore, 8 km, 16 km, and 24 km inland along M-45 for 0800 - 2400 EDT, July 6, 1967.

-19Li o Z LJ CD cr LAKE Ld a_:i6 0945-1005 EDT _ uJ 68 66,V L 0 1 2 4 6 8 10 12 14 16 18 20 DISTANCE IN KM FIGURE 11. Temperature trace obtained along M-45 in the morning, 0945 - 1005 EDT, July 6, 1967. 76 o0 74 LAKE__ _ _ 74 LLJ 72 L70 i 68 i t1, 1220-1240 EDT 0 1 2 4 6 8 10 12 14 16 18 20 DISTANCE IN KM FIGURE 12. Temperature trace obtained along M-45 in early afternoon, 1220 - 1240 EDT, July 6, 1967.

-20started at 1415 EDT when the lake breeze had already advanced inland, indicated that the lake effect reached 3 km inland and showed a 4~F km-1 gradient near the shore, Figure 13. By about 1500 EDT, the lake effect reached 4 to 6 km inland, Figure 14. At this time the westerly component of the lake breeze was 2 to 3 m sec1 at the shore. Thus rapid modification of the lake air took place as it traveled over land. By 1600 EDT, the temperature gradient reached 6 km inland, Figure 15 and 16. The strongest gradient was still present near the shore. At 1730 EDT, the gradient reached 8 km inland and at the same time the wind shifted to westerly at the 8 km station. In early evening the lake breeze had penetrated to about 10 km inland Figure 21 and 22. WJBL station, 10 km inland and 17 km south of M-45, had a wind shift to westerly at 2020 EDT. The sharp temperature jump 9 km inland at 2000 EDT, Figure 2C, is difficult to account for, but may be due to local effects. A summary of the data is plotted in Figure 23, which shows the variation in time and space of the temperature field. In the morning, the temperature gradient was absent as uniform heating took place everywhere. At 1300 EDT, when the lake breeze started, the temperature dropped slightly near the shore. As time went by, the temperature leveled off at points progressively further inland reaching 8 km inland by 1730 EDT. From then on, the progress of the front was difficult to follow as the temperature started to

-2176 LAKE __ __LAKE I0 0 74:D / Ld 0 —: 1415 - 1435 EDT, July 6, 1967. u, L AKE ____LAKE 68 I 1415-1425 EDT IZ;;25-1435 EDT 66 0 1 2 4 6 8 8 6 4 210 DISTANCE IN KM FIGURE 13. Temperature trace of the lake breeze, 1415 - 1435 EDT, July 6, 1967. LAKE LAKE 76 — 72 72 70i w 71 " 68 1435-1450 EDT 1450-1502 EDT 66... — 0 12 4 6 8 10 8 6 4 2 0 DISTANCE IN KM FIGURE 14. Temperature trace for 1435 - 1502 EDT, July 6, 1967.

-2276 LL 0 LAKE _ 2_ __ 1___ ___2 z 74 X =: 72 - a-/ S 70w 68 - - 1526-1545 EDT 66 0 2 4 6 8 10 12 14 16 1820 22 DISTANCE IN KM FIGURE 15. Temperature trace for 1526 - 1545 EDT, July 6, 1967. LAKE 76 V 2 74 2_' UJ FIGUE -1545- 1610 EDT 66 16

-2378 L AKE____ E 0 1762 4 6811214 86 202 LLJ 6, 1 1617 -1640 EDT...._____ 1 1 L..J.. __ 0 I 2 4 6 8 10 12 14 16 18 20 22 DISTANCE IN KM FIGURE 17. Temperature trace for 1617 - 1640 EDT, July 6, 1967. ~]-,.A_-LAKE 7-........ r-_ I 70 1 —-— _01__D_ 66 ------ July 6, 1967.

-2476 78 _..... 74 72e 70 -- - 68 1710-1730 EDT 66 0 I 2 4 6 8 10 1214 16 18 20 DISTANCE IN KM FIGURE 19. Temperature trace for 1710 - 1730 EDT, July 6, 1967. AA ---— ^ ^^^ —-— LAKE LL ^ 72 Z 70 68___ 1730-1750 EDT 66 — 20 18 16 14 12 10 8 6 4 21 0 FIGURE 20. Temperature trace for 1730 - 1750 EDT, July 6, 1967.

-2578 IL 0 1838-1855 EDT 76 LAKE z 74 72 0: j 1957-201855 EDT 66 0 12 4 6 8 10 12 14 16 DISTANCE IN KM FIGURE 22. Temperature trace for 1957 - 2015, Juearly evening, 1838 1855 EDT, July 6, 1967.

72 73 16968 67 68 69 7071 72 7374 75 76 767 69 6 TIME EDT JL \.' 19/66 II-;,, / ^A j \ / ~12 * ** *1 * ~y *fi ** * * 77 f II * I ** * * * f 1(1(i e W t~ l /* * I I0 9 \ / 75jl\ 8 z 7- I 6 -j z 766 u - IY74 W 5 Z [ 4 | / j — — \) rl/\\\1T66 2^ * ZkI 6 I 10 II 12 13 14 15 16 17 18 19 20 21 TIME, EDT JULY 6, 1967 FIGURE 23. Temperature distribution on the lake breeze day of July 6, 1967. Temperature are in OF. Dashed line shows position of the lake breeze front.

-27drop everywhere due to decreased insolation. As gradients formed in the afternoon, the isotherms in Figure 23 oscillated in small waves. This may have been either due to inaccuracies in the sensor or to a pulsating behavior of the lake breeze front. It has been suggested by Moroz (1965) that the lake breeze front penetrates at an irregular rate. In this manner the lake air sweeps past a certain point, stalls so that warming takes place before a fresh surge of cool lake air sweeps through. The progress of the lake breeze could be followed well by this method; the temperature changes seem to agree with the times of the wind shifts at the 8 km station, the WJBL tower, and the USWB station at Muskegon Ariporto Modification of air was quite rapid and when the front was well inland, the temperature gradient across it was quite weak. 3.3 Results of July 8, 1967; Weak Southwesterly Gradient Flowo On July 8, winds were light between south and southwest as a high pressure center was located over the east coast and a trough was approaching from the west. The local situation is represented in Figure 24. The prevailing wind at the stations along M-45, Figure 25, was southwesterly at 3-4 m sec with a westerly component of 2 - 3 m sec the shoreline station, however, had a predominantly southerly flow with only some slight westerly component in the afternoon. The sky was mostly clear, with only some light cumulus clouds present.

-28MKG 64 MtE M-45 P-RR LAKE MICHIGAN WJBL 77AZO soORD 174I 80 I74 / MDW / 7SBN,s a_1P 1200 EDT, JULY 8, 1967 FIGURE 24. Mesoscale surface winds and temperatures around the southern basin of Lake Michigan at 1200 EDT, July 8, 1967. Two runs were made that day: one around noon and one in late afternoon. The run of 1130 - 1203 EDT is plotted in Figure 26. Temperatures inland reached 78~F while at the shore it was 72~F. A gradient of 1.5~F km1 was found the first 2 km inland. The gradient got progressively weaker further inland and the temperature leveled off at 8 km. The plot for 1810-1845 EDT, Figure 26, shows a temperature of 75~F at the shore and 83~F inland. The profile looks quite similar to the one at noon. The temperature gradient was again 1.5~F km1 near the shore, and the temperature leveled off 14 km inland.

-29SHORELINE 8 _6 4 - \ _ _ _ — 10 12 14 16 18 20 6 8KM 4, 2- - I I I I I I 10 12 14 16 18 20 z Z 6 16 KM 4_ 2 C 10 12 14 16 18 20 6 - 24 KM 42 I I I -ft.I I 10 12 14 16 18 20 TIME, EDT JULY 8, 1967 FIGURE 25. Westerly (solid line) and southerly (dashed line) components of the wind at the shore, 8 km, 16 km, and 24 km inland along M-45 for 1000 - 2000 EDT, July 8, 1967. The plot has been smoothed.

84 82 -_.... - 8 2 80 _ - t -- Z 78 Q- / ___ i w 76/ o - / --- 1130-1203 EDT ~74 ~~/ -4 ~ — 1810-1845 EDT 72 JULY 8, 1967 0 2 4 6 8 10 12 14 16 18 20 22 24 26 DISTANCE FROM SHORELINE IN KM FIGURE 26. Temperature profiles along M-45, at 1340 - 1203 EDT and 1810 -1845 EDT, July 8, 1967.

3.4 Results of July 16, 1967; Weak Southwesterly Gradient Flow. The situation was quite similar to that of July 8 as a light southwesterly wind prevailed, Figure 27. The westerly component of the winds inland was 2 - 4 m sec-1 while at the shoreline southerly winds prevailed, Figure 28. The winds from the 8 km station were missing. Fair weather prevailed all day. Again two runs were made: one around noon and one in late afternoon. The results are plotted in Figure 29. Between 1120 and 1200 EDT, temperatures ranged from 64~F at the shore to 69~F further inland. The gradient was strongest at the shore, about 1.2~F km, and ceased 10 km inland. Between 1740 and 1820 EDT, the corresponding temperatures were 69~F and 74~F. A gradient of 1.0~F km-1 existed near the shore and the temperature leveled off 12 km inland. -31

-32( \?2 MKG 74 MKE20 \ M-452 GRR 5 \ LAKE MICHIGAN WJBL AZO \ /?70,A BEM 52 72 ORD 214 72MDV CGX 56 69 SBN 75 JOT 53-1/ ~5S I3?2Sj^1~ ~1200 EDT, JULY 16, 1967 FIGURE 27. Mesoscale surface winds and temperatures around the southern basin of Lake Michigan at 1200 EDT, July 16, 1967. 6 - 4- 2_ SHORELINE 0 — I I I I I - 08 10 12 14 16 18 4 2 2 I ------ - 16 KM _ o- I I I^1 24 KM 08 10 12 14 16 18 3 4 08 10 12 14 16 18 TIME, EDT JULY 16, 1967 FIGURE 28. Westerly (solid line) and southerly (dashed line) components of the wind at the shore, 16 km, and 24 km inland along M-45 for 0800 - 1800 EDT, July 16, 1967. The plot has been smoothed.

-3370O.L 68 - z_ i 66 — _ j/i~ a/~ — 1120-1140 EDT -— 1200-1140 EDT 64 _ l 1i I I I I I I I I I I I 0 2 4 6 8 10 12 14 15 13 20 22 24 DISTANCE FROM SHORE IN KM 74 Uz 72 12 / 1740-1755 EDT 70- / 6 8 I I I I I I I I I I t 0 2 4 6 8 10 12 14 16 18 20 22 24 DISTANCE FROM SHORE IN KM FIGURE 29. Temperature profiles along M-45, at 1120 - 1200 EDT and 1740 - 1755 EDT, July 16, 1967.

4. SUMMARY AND CONCLUSIONS The results did show different temperature distributions under several different wind situations. With moderate westerly winds, the lake influence extended furthest inland and the temperature increased quite uniformly from the shore to about 10 km inland, beyond. which the gradient became small. VWith lig:t_ southwesterly winds, a noticable lake influence did not reach as far inland, and a relatively strong temperature gradient was confined to within a few km from shore. With a lake breeze during an easterly gradient wind, the temperature gradient changed rather abruptly inland at the lake breeze front; and the strongest gradient was observed within 2 km from the shore. Thus the inland penetration of the lake effect seems to be a function of the strength of the onshore component of the wind. When a lake breeze had penetrated some distance inland, the air right behind the front was only slightly cooler than the air ahead of it. At a particular location inland, the passage of the lake breeze front caused the air temperature to level off rather than drop. Only at and near the shoreline was the temperature observed to drop as the lake breeze front passed after which it held steady. -34

-35The position of the lake breeze front on July 6 as estimated from the temperature traces agreed quite well with the observed wind shifts at the stations. However, this estimation is accurate to only within 1 or 2 km for several reasons. First, the accuracy of our system is not precisely known. Second, though the response of the system has been measured in the laboratory, it is hard to judge how it actually responds in the field where rapid, small scale oscillations in temperature act on the system. Third, many local differences in the nature of the underlying surface cause irregularities which in different degrees mask temperature differences. For example, a wooded area and an open field would have different effects on the temperature and the wind. Finally, the boundary between the lake air and the land air becomes less pronounced further inland because of the modification of the lake air and mixing of air across the lake breeze front.

REFERENCES Bierly, E.W., 1968. An Investigation of Atmospheric discontinuities induced by a lake breeze. Ph.D. thesis, Dept. of Meteorology and Oceanography, Univ. of Michigan, Ann Arbor, Mich., 150 pp. Daniels, P.A., 1965. The urban heat island and air pollution. Master's thesis. University of Alberta, 144 pp. Duckworth, F.S. and J.S. Sandberg, 1954. The effect of cities upon horizontal and vertical temperature gradient. Bull. Amer. Met. Soc., 35, 198-207. Hewson, E.W. and L.E. Olsson, 1967. Lake effects on air pollution dispersion. J. Air Poll. Cont. Assoc., 17, 757-761. Lyons, W.A. and J.W. Wilson, 1968. The control of Summertime cumuli and thunderstorms by Lake Michigan during non-lake breeze conditions. SMRP paper No. 74. Department of Geophysical Sciences, The University of Chicago, 32 pp. Moroz, W.J., 1965. The lake breeze circulation along the shoreline of a large lake. Ph.D. thesis, Dept. of Meteorology and Oceanography, Univ. of Michigan, Ann Arbor, 120 pp. -36

-37Olsson, L.E., A.L. Cole, and E.W. Hewson, 1968. Observed land and lake breeze. circulation on the eastern shore of Lake Michigan, 25 June, 1965. Dept. of Meteorology and Oceanography, Univ. of Michigan, Ann Arbor, 93 pp. Summers, P.W., 1965. An urban heat island model, its role in air pollution problems, with applications to Montreal. First Canadian Conference on Micrometeorology, Toronto. 0 Sundberg, A., 1950. Local climatological studies of the temperature conditions in an urban area. Tellus, 2 (3), 221-231.

UNIVERSITY OF MICHIGAN 3 9015 03026 9974III 3 9015 03026 9974