Loading...
HomeMy WebLinkAboutSolar Radiation Data For Kodiak, Alaska, May 1984 RECEIVED JuL 2 4 1984 SOLAR RADIATION DATA FOR KODIAK, ALASKA by Gerd Wendler Geophysical Institute, University of Alaska Fairbanks, Alaska 99701 UAG R-299 May 1984 LIBRARY CoP’ SOLAR RADIATION DATA FOR KODIAK, ALASKA Gerd Wendler Geophysical Institute University of Alaska Fairbanks, Alaska 99701 ABSTRACT Solar radiation data from Kodiak (57°45'N, 152°30'W) which is located in the maritime climatic zone of Alaska were collected from January 1980 to May 1982 and analyzed. These are the first data of this kind ever taken at Kodiak. Global radiation with an annual mean value of 108 Wm-2 is low, due to the high amount of cloudiness, which is typical of the maritime climate of Alaska. In summer a horizontal surface receives more solar radiation than a south slope inclined at an angle equivalent to the latitude; in spring and autumn the differences are minor, while in winter the south slope receives about twice the radiation of the horizontal surface. Relationships were established between the cloudiness and global radiation for Kodiak. This relationship was tested for Annette, also located in southern Alaska, for which station historical radiation data exist. Good agreement was found, which allows one to calculate the solar radiation for other stations in southern Alaska, for which no insolation data have ever been measured. INTRODUCTION Radiation data for Alaska are sparse (Wise 1980), and the few data which exist are not always trustworthy due to the infrequent calibration. As part of a national effort with funding from the Department of Energy, Kodiak was chosen to become one of the four newly established radiation stations in Alaska. Prior to this time no radiation data had ever been collected in Kodiak. This paper analyzes the data, which are not only of scientific interest, but also might prove of practical value, as they can be used as a base for solar radiation assessment of alternative energy sources. Calculation of solar energy collectors, photovoltaic arrays, or size and direction of windows in energy efficient architectural designs of buildings might be aided. AREA, INSTRUMENTATION AND TIME OF OBSERVATIONS Kodiak is situated near sea level on Kodiak Island at 57°45'N, 152°30'W in southern Alaska (Figure 1). The area is situated in the maritime climatic zone of Alaska (Searby 1969), with cool summers and relatively warm winters. The warmest month is August (54.9°F), and the coldest December (29.9°F), while the mean annual temperature is 40.7°F. The precipitation is 56.7 inches, about 4 times as high as Anchorage, nearly 5 times as high as Fairbanks, and an even higher ratio for Barrow. Anchorage, Fairbanks and Barrow are the other three stations where radiation equipment is installed (Becker and Leslie 1983, Wendler and Eaton 1983). At Kodiak the cloudiness is high with a mean annual value of 74%, which is very similar to the value found for Barrow. The large ratio of precipitation therefore indicates that the clouds of Kodiak are much "thicker", holding more water droplets, which have also an effect on the radiative characteristics of the clouds. In the fall of 1979 a Kipp and Zonen pyranometer (No. 773918) measuring solar radiation on a horizontal surface (Figure 2), and a Lambda cell measuring solar radiation on a 57° inclined south slope were installed on the roof of the building which houses the National Weather Service, which is situated near to sea level. The instrumentation was calibrated against a "standard" Eppley PSP, which was in turn calibrated against the national standard in Boulder, Colorado (Flowers 1974). Also calibration was carried out against a Linke Feussner actinometer, using the shading method. The latter instrument was calibrated in Boulder as well as at the World Radiation Center in Davos, Switzerland. The accuracy of the Kipp and Zonen Pyranometer is + 2%. Global radiation (H) is the sum of direct solar radiation and the diffuse sky radiation, which is the energy a flat plate collector receives. The energy received on a south slope is normally larger on clear days than that on a horizontal surface. Furthermore with the low solar elevations (in winter), over most of Alaska, these values are more useful for energy application. However, only the potential solar energy is evaluated; that is the energy which is available from the sun and sky. Due to losses which will occur in any system, the energy available from, for example, a flat plate collector, will be less (e.g. R. Seifert 1981). Depending on the design of the system the losses range widely and may be substantial. The efficiency of a photovoltaic system (the direct conversion of solar to electrical energy) may be as low as 10%. The data were recorded on a Speedomax strip chart recorder. Besides the output of the instruments, the zero point was constantly recorded. The Measurements were scaled at hourly intervals and put on the VAX computer of the Geophysical Institute. During the first months of operations, some problems were experienced, hence our analyses started only on 1 January 1980, from which date on missing data are infrequent. In May 1982 the measurements were discontinued. The roof of the building, on which our equipment was located, needed repair, hence all instrumentation had to be removed. Therefore, a time interval of 2 years and 4 months is available for analysis of the global radiation data. For the radiation received on the inclined south slope only data for the first 8 months proved to be reliable. It appears, that after that time the sensor was not correctly aligned. RESULTS In Figure 3 a time series of the solar radiation for the 28-month period is given. The typical annual course, with low values in winter at lower solar elevations and higher values in summer, can be seen. The mean flux is 108 Wm-2, a relatively low value for the latitude of 57°N (Kondratiev 1969). The value is quite similar to Anchorage which has a value of 107 wm-2 (Becker 1981), Fairbanks at 114 Wm-2 (Wise 1979) or for a later period 103 wm-2 (Wendler and Eaton 1983) and Barrow which shows a value of 103 Wm-2 (Wise 1979). The relative low value for Kodiak, taking into account its lower latitude and expected higher global radiation, is believed to be caused by the high amount of cloudiness in Kodiak. Further, the clouds are normally optically much thicker than those found in Barrow, as they contain many more water droplets. The measured value is also in excellent agreement with the one calculated for Kodiak (105 Wm-2) (Knapp et. al. 1980). Figure 4 displays the annual course of the solar height at noon. In winter, values around 10° are observed, while in summer values rise to the upper fifties. This is, of course, in great contrast to Barrow, which is situated north of the Arctic Circle, and does not receive any sunshine at all during midwinter. Figure 5 shows the diurnal course of the solar elevation for midwinter, midsummer and the equinoxes for Kodiak. This graph shows that not only the solar height at noon changes during the year, but that also the length of the day changes drastically, a fact of which we are all aware. Figure 6 shows the global radiation of Kodiak as point values. While midwinter values are normally below 30 wm-2, maxima in summer are often more than 10 times as high. The encompassing curve shows the maximal possible value, which is the extra-terrestrial (ET) radiation on a horizontal surface for the latitude of Kodiak. This value would be recorded if there would be no atmosphere. Even on a cloudless day, maxima reach only about 80% of that curve, as absorption by the heavy three-atom gases (water vapor, C09 and 03) in the atmosphere and the aerosols reduce this theoretical value. Further, the graph indicates that maxima in spring are nearer to the ET values than the fall values. This is believed to be due to the smaller amount of water vapor in the atmosphere in spring after the cool winter when compared to autumn. Table 1 gives the insolation values on a monthly basis. Besides the mean and the extremes, global radiation for different classes of cloudiness are also presented. It shows, for example, that the amount of radiation received on a clear day is more than twice the amount received on a cloudy day. The relatively high value for April is being caused by a below normal cloudiness for the years of observations, as well as the fact that frequently the clouds observed during this month were cirrus; these thin clouds have little effect on the amount of radiation received. In their classical paper Liu and Jordan (1960) related different 4 radiation fluxes to each other. Among their findings was the Kt, which is the ratio of global to extra-terrestrial radiation, and shows a linear de- pendency to the percentage of time with direct solar radiation. A fairly linear relationship was found, as long as monthly means were used. This is similar to the original work of Angstrom (1924), who related the global radiation to the sunshine duration and used the latter to estimate the global radiation for other stations, where no global radiation was measured. We wanted to test this relationship for Kodiak to see if a sound relationship could be established for calculating the global radiation for other places in southern Alaska. The global radiation was measured and the extra-terrestrial radiation (ET) on a horizontal surface can be calculated, taking latitude and sun-earth distance into account. A monthly listing of ET for the latitude of Kodiak as well as for other latitudes of southern Alaska is given in Table 2. Sunshine duration was not measured for Kodiak, however cloudiness was obtainable. We used the data between sunrise and sunset. As there is a relationship between sunshine duration and cloudiness, this substitution is permissible, but the relationship is expected not to be quite as good, as no distinction between different cloud types has been made. In Figure 7 the relationship between KT = H/ET is plotted against cloudiness. The best linear fit is put through the monthly data points. The strongest deviation from a linear relationship is being found for 0 and 10/10 cloudiness. For 0/10, which is totally clear skies between sunrise and sunset, a Kt of 0.71 is found while the best fit would result in a value of 0.75. The observed value of Ky = 0.71 means that on the average over the year on a clear day 71% of the extra terrestrial radiation reaches the earth's surface. The rest is being absorbed in the atmosphere or reflected 5 to space. This value fits well with observations of other authors at other places on earth. In England, values between 0.73 and 0.81 were reported by Hanna and Sicon (1981), while Glower and McCulloch (1958) found values between 0.74 and 0.84 for the tropics and the subtropics. Rao et. al. (1983) reported values around 0.76 for Oregon. The value we found is slightly lower than all the other measurements, however, our station has a higher latitude, with the longest average paths through the atmosphere. With increasing path length of the solar beam absorption becomes increasingly important. Glower and McCulloch suggested a relationship which takes the latitude into account. This relationship would result for clear sky conditions for Kodiak in a value of 0.68 compared with our measured value of 0.71. The influence of the pathlength on Ky for clear sky conditions can also be seen in the annual cycle of Kt. In winter, lower values are found (mean of November, December, January = 0.65), while in the summer (mean of May, June, July = 0.75) less absorption takes place in the atmosphere. Also, in spring higher values were observed than in fall for the same solar elevation. This cannot be due to different lengths of the solar path, but is believed to be caused by the lower amount of water vapor in the atmosphere in spring when compared to fall. Water vapor is one of the important absorbers in the atmosphere. For overcast skies (10/10 cloudiness), a mean annual value of KT = 0.23 is being observed, which means that about 3/4 of the extra terrestrial radiation is reflected to space or absorbed within the atmosphere. Again this value is in agreement with findings for other locations. The above mentioned authors found values ranging from 0.18 to 0.30. Our data for overcast conditions show that the values in spring are somewhat higher than those found in fall for the same solar elevations. This is believed to be due to multiple reflections. 6 In spring, multiple reflections occur between the snow-covered ground and the clouds, which increases the amount of incoming global radiation (Wendler et. al., 1981), while in fall the surface is bare, hence the reflectivity of the ground is low. The established relationship is not only of scientific interest, but also has practical importance, as it allows the calculation of the amount of global radiation for stations in the same climatic region, where no radiation measurements have been carried out. This was also the motivation for Angstrom (1924) who 60 years ago tried to establish this for the first time. The linear relationship we established between KT and cloudiness, can explain most of the variance. For testing the relationship between cloudiness and solar radiation (the good relationship for Kodiak is shown in Figure 8), we looked for another station. The only other station with solar radiation data in the maritime climatic zone of Alaska is Annette, located in Southeastern Alaska at 55°02'N and 131°34'W. There global radiation measurements were carried out between July 1952 to December 1975 (Searby 1968, Wise 1979). Using the relationship from Kodiak Ky = 0.753 - 0.041C with Ky = H/ET and C the cloudiness in tenths, the radiation was calculated for Annette and is plotted against the actually observed values. Good agreement was found (Figure 9) with a correlation factor of 0.995 between measured and calculated values. This result indicates that the global radiation for other places in southern Alaska can be calculated. To facilitate this task, in Table 2 the extra-terrestial radiation on a horizontal surface is tabulated for monthly sums for the different geographic latitudes of southern Alaska. The global radiation is then: H = ET (0.753 - 0.041C) 7 with C the cloudiness in tenths and ET being obtained from the table. Diurnal Variation In Figure 10 the mean diurnal variation of the global radiation is given for December, March and June. September has a very similar course to the one found for March. As expected, not only the day length increases when going from midwinter to midsummer, but also the amount of energy received increases steadily. While in December the radiative flux at noon is less than 100 Wm-2 for an "average" day, it increases to 360 Wm-2 in March and 600 Wm-2 in June. Inclined Surface A south slope inclined to the elevation of latitude, in our case 57°, receives, in general, more radiation when the sun is shining, but less during times with total cloud cover and only diffuse radiation, as the sensor does not see the whole sky. Also in winter, the differences for a clear day are expected to be larger than in summer. The path of the sun is long in summer , and no direct solar radiation falls on the inclined south slope during the early morning and late evening hours. This reduces the advantages of a south slope. The amount of cloudiness is high for Kodiak (74% annual mean), hence the advantages of a south slope over the horizontal surface will not be as pronounced as for a station where more sunshine is being observed. In Figure 11 the ratio of radiation received on an inclined to that on a horizontal surface (global radiation) is given for 0-5 tenths cloudiness as a function of the solar height at noon. Values around 10° solar elevation represent midwinter, the upper 50° are midsummer values. It can be seen that in midwinter the south slope receives up to 5 times as much radiation as the horizontal surface, while in summer the advantages of a south slope vanishes. Now the increased radiation during the noon hours are compensated 8 by less radiation during the early morning and late evening hours. For overcast conditions, the majority of all days receives less radiation on the south slope than on a horizontal surface. This is observed during the whole year (Figure 12). The few days when the south slope is receiving more radiation are days with thin cloud cover, probably cirrus or cirrostratus, through which the sun is penetrating. Regarding the monthly means, the south slope holds substantial advantages over the horizontal surface in winter; about twice as much solar radiation is received (mean of January and February), while in summer the inclined surface receives about 20% less radiation than the horizontal surface (mean of May, June and July). During spring and autumn no pronounced differences are observed. The annual sums are similar, as the large advantages of the south slope in winter occur at a time when the solar radiation is weak and are hence compensated by the smaller disadvantages of the south slope in summer, when radiation is at its maximum. Advantages of the south slope over the horizontal surface in the annual sum were observed in Anchorage and Fairbanks (Wise 1980, Wendler and Eaton 1983). However, the amount of cloudiness for these stations is somewhat lower, which enhances the advantages of the south slope. Further, the clouds are believed to be optically thicker at Kodiak, an indirect indication for this is the higher amount of precipitation. ACKNOWLEDGEMENT The work was supported by DOE grant No. EY-77-G-06-1060, the Alaska Council for Science and Technology, Grant. No. W81-01 and State of Alaska funds. The U. S. Weather Service, Kodiak under the supervision of Mr. Sam Welch, Meteorologist in Charge, took care of the instrumentation. Mrs. E. Curry reduced the data, Mr. Y. Kodama did the programming, and Ms. M. Hall and Dr. F. Eaton helped with other aspects of the investigation. To all of them my sincere thanks. 10 REFERENCES Angstrom, A., 1924. Solar and terrestrial radiation. Quart. J.R. Met. Soc. 50, 121. iy Becker, R., 1981. Comparison of solar radiation measurements on a horizontal inclined, and vertical surface in Anchorage, Alaska. AEIDC Publication, 26 p. Becker, R. and L. D. Leslie, 1983. Solar radiation assessment for Anchorage, Alaska. AEIDC Publication. 35 p. Flowers, E. C., 1974. Atmospheric turbidity measurements with the dual wavelength sun photometer, WMO Publication No. 368, Geneva, Switzerland. Glover, J. and J.S.F. McCulloch, 1958. The empirical relation between solar radiation and hours of sunshine. Quart. J.R. Met. Soc., 84, 172. Glover, J. and J.S.G. McCulloch, 1958. The empirical relation between solar radiation and hours of bright sunshine in the high-altitude tropics. Quart. J. R. Met. Soc., 56. Hanna, L. W. and N. Siam, 1981. The empirical relation between sunshine and radiation and its use in estimating evaporation in northeast England. J. Climatology, 1, 11. Knapp, L., T. L. Stoffel and S. D. Whitaker, 1980. Insolation Data Manual. SERI/SP 755-789, 282 pp. Kondratyev, K. -Y., 1969. Radiation in the atmosphere. Academic Press, New York, 912 pp. Liu, B.Y.H. and R. C. Jordan, 1960. The interrelationship and characteristic distribution of direct, diffuse and total solar radiation. Solar Energy, 4, 1. Rao, C. R. , W. A. Bradley and T. Y. Lee, 1984. Some Comments on Angstrom Type Regression Models for the Estimation of the Daily Global Solar Irradiation, in press, Solar Energy. Searby, H. W., 1968. Climates of the States, Alaska, U. S. Department of Commerce Climatography of the United States No. 60-49. Seifert, R., 1981. A solar design manual for Alaska, University of Alaska, Institute of Water Resources, 163 p. Wendler, G. and F. Eaton, 1983. Solar Radiation Data for Fairbanks, Geophysical Institute Report, 58 p. Wendler, G., F.D. Eaton and T. Ohtake, 1981. Multiple reflection effects on irradiance in the presence of arctic stratus clouds. Journal Geophysical Research Vol. 86, No. C3, pp. 2049-2057. Wise, L., 1980. Analysis of Solar Radiation Measurements on an inclined surface in Anchorage, Alaska, AEIDC Publication, second printing, 20 p. a} 1 WwW m-2 u 1.0 0.2388 0.05687 0.00528 0.320 7.68 234 APPENDIX CONVERSION FACTORS Joule sec-!m-2 cal sec~1m2 BTU min-1m-2 BTU min-Lft-2 BTU hr-LFt-2 BTU day-!ft-2 BTU month-lFft-2 12 TABLE 1 Global radiation (Wm-2) received in Kodiak as function of cloudiness on a mean monthly basis. MONTH wo @ 10 i 12 AVE. 17. 52. 100. 164. 163. 232. 201. o nv Ga A NS 156.2 98. 66. 26. 15. Cloudiness in Tenths MAX. MIN. 0-1 2-5 6-9 10 49.2 2.2 30.5 29.7 22.8 8.6 111.5 4. 89.9 77.6 49.2 25.5 212.8 13.7 159.5 152.6 104.2 45.1 283.2 21.3 235.1 227.0 171.3 118.2 335.5 30.4 312.5 294.1 217.1 98.1 353.1 84.0 342.7 309.1 21950 “139.9 392.2 49.5 359.6 295.3 218.7 140.4 311.0 43.7 252.8 231.7 177.4 94.9 219.1 12.7 215.9 167.2 98.8 62.9 138.2 5.1 108.7 97.5 66.6 25.5 74.8 3.9 45.0 41.0 28.9 13.6 31.3 1.3 24.4 1.1 16.2 7. 13 TABLE 2 Extra-terrestrial solar radiation (Wm-2) on the horizontal and other latitudes in southern Alaska on a monthly basis. surface for Kodiak MONTH LATITUDE 55.0° 57.5° 60.0° 62.5° 1 72.5 56.2 40.7 26.5 2 136.3 118.6 100.9 83.4 3 237.5 221.1 204.3 187.1 4 350.0 338.3 326.2 313.8 5 440.5 435.0 429.4 423.9 6 482.3 480.7 479.1 478.3 7 460.6 457.1 453.6 450.5 8 383.3 374.1 364.5 354.7 9 277.5 262.7 247.5 231.9 10 170.4 153.0 135.4 117.7 nN 90.6 73.7 57.3 41.8 12 56.9 41.5 27.2 14.6 14 708 180° 170° 160° 150° ~ 140° 130° 120° 70° BARROW FAIRBANKS ‘| e \ ALASKA \ ANCHORAGE \ 6 9 oS KODIAK Slay . wert 170° 160° 150° 140° 130° Figure 1. Location map of Alaska. Figure 2. A Kipp and Zonen Pyranometer. osét 4c qNOSUYUrrfrwWUN 186 te 286T OW 3S (ON 0-S26 fC on OWS ul s S ir" Figure 3. GLOBAL RADIATION <W/M¥*2) OW Ol = SS ul Ss ou S ol S S Ss S S Ss i] A time series of the insolation received at Kodiak, January 1980 to April 1982. SOLAR HEIGHT AT NOON ¢DEGREES> 68 58 40 38 28 10 Figure 4. Annual course of maximum solar height for Kodiak. 68 ul © f s SOLAR HEIGHT ¢DEGREES)> nN i So © 12 B 3 6 9 12 15 18 21 24 HOUR Figure 5. Diurnal course of the solar path for midwinter, the equinoxes and midsummer for the latitude of Kodiak. 52a 408 S & © © RADI ATI ON INTENSITY (W/M¥*2) © 8 Figure 6. Global radiation data for Kodiak are given as point values. The encompassing curve is the extra-terrestrial radiation for the latitude of Kodiak, that is, the value Kodiak would be receiving if there would be no atmosphere. 1.8 Y=-@.O41X+ 0.753 ». R**2= O. 880 com me coef ofre aw: cece cenetfoendow wee cere see @ 1 2 3 4 5 6 c 8 9 18 CLOUDINESS (TENTHS) Figure 7. The ratio of global and extra-terrestrial radiation is plotted against cloudiness for Kodiak, Southern Alaska. 388 259 - n + N + $200 4 ‘ a N + + = + = + f15e—4 * > + ~ oO a + *, ©1288 — + + te + 58 — + + me ra : Sa EET Taam a 52 182 15a 289 258 380 CALCULATED (W/M**2> Figure 8. The observed global radiation is plotted against calculated values for Kodiak, Southern Alaska. 328 2598 288 15 OBSERVED (W/M**2)> Figure 9. 58 188 152 288 258 382 CALCULATED (W/M*¥*2)> The observed global radiation of Annette is plotted against calculated values, using the relationship found for Kodiak. A correlation factor of r = 0.995 was found. 622 ul i) ® f£ 8 ®& 3280 GLOBAL RADIATION ¢W/M**2) N g © 188 Q 3 6 9 12 15 18 21 24 HOUR Figure 10. Diurnal course of the observed global radiation for December, March, and June, Kodiak, Southern Alaska. RATIOCINC/GLO> Figure 11. 18 22 38 42 32 68 SUN ELEVATION (DEGREES? The ratio of radiation on an inclined surface to that on a hori- zontal surface is plotted against solar elevation for 0 to 5 tenths cloudiness. Note the increase of the ratio with decreasing solar elevations.