908 ' MONTHLY WEATHER REVIEW VoLv-gzl05Continentality in the Texas Coastal Zone KENNETH H. JEHNAtmospheric Sden~ Group, Th~ University of Texas, Austin 78712(Manuscript received 2 March 1977, in final form 4 April 1977.)ABSTRACT An analysis is presented of continentality in the Texas coastal zone, based on standard climatologicaldata from the existing National Weather Service network. In spite of prevailing onshore winds throughoutthe year, Texas coastal zone stations (Gulf of Mexico to 150 mi inland) exhibit a substantial degree of continentallty by any of the usual definitions. Continentality by the Conrad definition ranged from approximately 25% in the Brownsville area to about 38% inland, with a mean value of approximately 33% forthe. 1941-70 period in the entire coastal zone. Additionally, some measurements are presented of horizontal gradienlis of air temperature in summer andwinter along a line perpendicula~ to the Gulf coast from near Port Aransas to near Sinton, Tex., approximately 30 mi inland. Mean daily temperature range and other statistical parameters for these stations underscored the continentality findings, and demonstrated substantial horizontal gradients of continentality in theimmediate coastal zone, with a val~e of contL,/entality of about 20% at the Port A..ransas beach. Overwatervalues of continentality from summarized ship reports off the Texas coast were of the order of 18-20%.1. Introduction The degree to which land areas influence climate isexpressed by the concept of -ontinentality, first statedby Zenker in the 1890's (Harm, 1908; Conrad, 1946).Numerous modifications of the basic formula have beenmade, but all, including Zenker's original formulation,show 'the index of continentality for a given location todepend on the annual range of temperature and thelatitude of the station. Inclusion of the latitudecounts for the observed increase of annual temperaturerange with latitude, and makes possible the comparisonof continentality index values over large areas of theglobe. The empirical formula for continentality due 'toConrad (1946) expresses the degree of laud influence inan index k =E1.7~l/sin(~+10-)qwhere A is the difference in mean temperature (-C) between the warmest and coldest months at the stationand tb is latitude of the station. The two empiricalconstants were chosen in such a way as to make theindex zero for the "most marine" location and 100 forthe "most continental" among the stations studied.Conrad introduced the correction factor of 10-latitude to avoid the difficulty of low-latitude determinations of continentality. Most distributions of continentality in the literature appear to ut'dize the Conradformulation, although numerous modifications havebeen proposed by others (Gorczynski, 1920; $pitaler,1922; Brunt, 1924; Johafisson, 1931; Raunio, 1948;Hela, 1953). Other techniques for the expression of continentaland/or ocean effects have also been proposed: Kerner,1905 (index of oceanicity); Dinies, 1932 (air massclimatology); Leighly, 1938 (extremes of annual temperature); Berg, 1940 (ratio of continental to maritimeak masses); Calef, 1950 (interdiurnal variability of temperature extremes); Court, 1951 (temperature frequendes); Prescott and Collins, 1951 (lag of temperature behind insolation); Craddock, 1964 (interannualvariability of monthly mean temperature); Bailey,1968 (hourly temperatures related to annual range);Polowchak and Panofsky, 1968 (spectrum of dailytemperatures). The point is that continentality is not a simple concept, and a variety of schemes has been proposed toexpress the basic idea. In any case, a scale of continentalky is useful in mapping the relative influence of landand water areas for the various climatological stationsworldwide, giving numerical measures of the land (orwater) effect. Coastlines everywhere juxtapose continental andmarine environments, accounting for such phenomenaas the land/sea breeze. It is no surprise, therefore, tofind that isopleths of continentality generally parallelthe continental margins [e.g., Trewartha (1961) shownin Fig. 1']. However, it is dear from an examination ofsuch distributions that continentality "spills over" toadjacent water areas, especially on the east (lee) coastsJULY1977 KENNETH H. JEHN 909~20Fro. 1. Isopleths of equal continentality for Anglo-America, using V. Conrad's formula (Trewartha, 1961).of continents in middle and high latitudes, and that suchsignificant water bodies as the U.S. Great Lakes introduce a marine effect sufficient to distort the continentality isopleths. In North America, only the west(windward) coast exhibits continentality isoplethsparallel to the coastline, as well as an index which approaches zero near the coast (see also MacKay and Cook,1963). Such regional features of continentality distribution have prompted special studies on a smaller scale:Leighly, 1941 (U. S. Great Lakes); Fobes, 1954 (NewEngland); D'Ooge, 1955 (western United States);DeBrichambaut and Wall~n, 1963 (Near East); Kopec,1965 (U. S. Great Lakes). A similar study is reportedhere for the Texas coastal zone.2. Texas coastal zone studya. Continentality distribution Continentality by the Conrad (1946) formula wasdetermined for 40. NOAA climatological stations in theTexas coastal zone for the standard period 1941-70,and for an additional 15 stations with near 30-yearrecords) Thus, 55 stations were utilized to yield continentality isopleths (Fig. 2) from the Gulf coast toapproximately 150 mi inland.2 ~ Bracketed data plotted in Fig. 2 are for stations with 28 or 29year records during 1941-70, or for stations with 30-year recordsoverlapping 1941-70 by 26-29 years. 2 "Miles" are statute miles. The trend of the isopleths in Fig. 2 is nearly parallelto the Gulf coast; however, the upper Texas coast ismore continental (~30%) than the lower coast(~25~Vo), and there are other departures from thesimple parallel mode suggested by the Trewarth~ distribution of Fig. 1. For instance, there is a tendency forsmaller continentality to appear in the Rio Grandevalley, compared to the continental (semi-arid) highlands just to the north. There is also a tendency for lesscontinental air to appear in the region between Cotullaand San Antonio, and for the urban complex along anapproximate Houston to Port Arthur axis to exhibitlarger values of continentality than the surroundings.The mean value of the continentality index for the 55stations for the 30-year period is 33.1%, with standarddeviation of 3.2. The continentality distribution shown in Fig. 2 mustbe considered preliminary, because no effort has beenmade to examine individual station exposures, nor tofind other sources of error. There is no assurance thatthe 30-year period is a stable climatological period;indeed, a study of the data showed that 5-year meansof continentality index generally increased slightly overthe network during the 30-year period. It is difficult totell whether this is a real increase in continentality inthe Texas coastal zone, or merely an increase in urbaneffects around the existing climatological stations.910 M O N T H L Y 'W E A T H E R R E V I E W VOLU~aE 105FjO~WCONTI NENTALITY ISOPLETHS,TEXAS COASTAL ZONEFIG. 2. Continentality isopleths, Texas coastal zone, Conrad definition, period 1941-70, 55 stations.b. ~ield measurements The small station density overland within 5-10 miof the Gulf Coast (Jehn, 1974, pp. 7-10) does not permita detailed examination of continentality by the Conradformula or similar expressions which depend on meanannum temperature in the region where the largestgradients would be expected. In a study designed to'monitor daily temperature range near the Texas coast,four hygrothermographs in standard instrument shelters were employed: one on the back beach on MustangIsland at Port Aransas, ~0.1 mi from the Gulf waterline; a second 1500 ft inland from the first at the University of Texas Marine Science Institute also onMustang Island; a thirda in Aransas Pass on the mainland, ~ 10 mi from the Gulf; and the fourth near theheadquarters building of the Welder Wildlife Refugenortheast of Sinton, Tex., ~30 mi from the 'Gulf. Thea Courtesy of Elgie Wingtield, Institute boat captain.hygrothermographs were operated from 19 December1974 to 25 January 1975, and from 6 June to 10 July1975, representing a winter and a summer period.(Difficulty with the humidity elements left the tempera-.tures as the only reliable data set during these twoperiods.) Hygrothermographs in standard shelters werechosen over other instrumentation so that comparisonof the measurements with standard climatological datawould be feasible. The diurnal variation of temperature at the~e fourstations was examined for the ,ummer and winterperiods and compared with the daily range over thesame'periods at NWS reporting stations from AransasPass inland to Kenedy. The comparative data aresummarized in Table 1. The continental effect is striking, and is evidentlyconcentrated in the first 30 mi or so from the Gulf ofMexico during the summer period, during which onlyone cold front passage was noted at Port AransasJULY 1977 KENNETH H. JEHNTABLE 1. Mean daily temperature range (-F) (MDR) from "Beach" Station to Kenedy, Tex.911Approximate distance from Gulf (mi)MDR Summer period (6 June to 10 July 1975)1 2 3 Aransas 4"Beach" "Institute" "Elgie" Pass "Welder" Sinton Beeville Kenedy 0.1 0.4 10 10 30 30 60 80 4.3 7.5 12.5 13.9' 17.6 19.0 18.8 20.6MDR Winter period (19 December 1974 to 2~ January 1975) 1 2 3 4"Beach" "Institute" "Elgie" Rockport !'Welder" Sinton Beeville Kenedy 9.8 11.2 18.7 17.6 21.2 22.7 23.5 25.0(11-12 June), although northeasterly winds were notedon other occasions as well. Similar results are noted forthe winter period, except of course that the mean dailytemperature range is larger, especially close to the coast,compared to the summer data. While these data exhibit various degrees of continentality, the findings are not easily compared to thecontinentality distributions based on annual range oftemperature. Bailey (1968) has examined the relationship of hourly temperatures to annual range, finding auseful correlation between annual range of temperature and the standard deviation of hourly temperaturesaround the annual mean for a large number of standardreporting stations (Fig. 3). The physical basis for thiscorrelation suggests that there might be a shorterperiod validity applicable to the data available in thepresent study. Accordingly, temperatures were readfrom the hygrothermograph charts at 2 h intervals andtabulated for each station. Means and standard deviations were calculated for the winter and summer periods,with the results shown in Table 2. The differences in arithmetic means for the four stations in one season are probably not signifi~ant, but inboth winter and summer the increasing standard deviation with increasing distance from the Gulf is indicativeof increased continentality. Plotting the standard deviations shown in Table 2versus the extreme range for the winter and summerperiods for the four stations (Fig. 4) yields a relationship similar to that of Bailey (1968), suggesting that afundamental correlation indeed exists between standarddeviation of short interval temperatures and the rangeof temperatures during the period in question. It is notpossible to interpret these results directly in terms ofbB.J~ 25r~t~o_,,=, 2o~Br~g ~s-i[1_o_c5 ~oa->hi 5r~r~<[r~Z<[~o3 5 I0 15 20 25 30 35 40 45 50 55 60 MEAN ANNUAL RANGE OF TEMPERATURE , Am (-F ~Fro. 3. Standard deviation of hourly temperatures vs mean annual range of temperature (after Bailey, 1968).912 MONTHLY WEATHER REVIEW VOLUI~IE10515(/3IO 5O SUMMER DATA 6 JUNE- I0 JULY 1975x WINTER DATA 19 DEC 1974-25 JAN 1975 . xJ ' I0 20 ~ 30 40 50 60 EXTREME RANGE (-F) Fro. 4. Standard deviation of 2 h temperatures vs extreme range, Texas coastal zone field measurements, 1974-75.continentality based on annual range. However, thestations within 30 mi of the Gulf of Mexico in Table 2could be interpreted4 as having a standard deviation ofhourly temperatures for the year of the order of 12-F.Bailey's analysis indicates that the annual range oftemperature corresponding to this standard deviationis approximately 25-F. An annual temperature rangeof this magnitude at the latitude of Port Aransas wouldyield a continentality index (Conrad) of close to 25%,a value somewhat less than the extrapolated value inFig. 2. Data from a single year (1974-75) can hardly becompared with a long-period record, but the order ofmagnitude of this continentality index, along the coastalbend seems reasonable. Judging by the smaller standard~teviations exhibited much closer to the coast in Table 2,it would appear that the continentality index at'!Beach" for 1974-75 could be on the order of 20O/o. 4 Bailey (1968) noted a relationship among dispersions aboutthe mean such that the sum of the variance of hourly temperatures about the monthly mean (~m) and the variance of themonthly mean temperature about the annual mean (*~mu) equalsthe variance of the hourly mean about the annual mean (~u).In the present case, ~h~ is approximated by the standard deviations shown in Table 2; (r,~ can be determined from climatologicaldata for such stations as Rockport and Sinton to be in the orderof 10-F. Thus, ~h~ for "Elgie" and "Welder" must be in the orderof 13-F, giving an annual range (A,~) of the order of 28-F (Fig. 3),and a continentality index of the order of 30%, in agreement withthe isopleths of Fig. 2. Since *am for "Beach" and "Institute" aresmaller than those for Elgie and Welder, an2m for the year can beestimated to be in the order of 35. If a~mu is taken to be about 100,tr~v will be of the order of 12-F.c. Overwater values Suitable temperature data from overwater locationsare unavailable in the sense of long records at a singlelocation. However, weather observations are maderoutinely (and with a fair-weather bias) by ships whichoperate in the Gblf. The National Climatic Center,under the direction of the U. S. Naval Weather ServiceCommand (1975), has summarized meteorological observations for overwater locations in a series of publications. The data so summarized for limited overwaterareas are applied to the "centroid" of the area. Thelocation for summarized:ship reports in the extremewestern Gulf' of Mexico (Area 29, Corpus Christi) is27.5-N, 95.8-W, and the primary period of record is1952-71. Utilizing the frequency distributions of temperatures at the synoptic hours for ships in this area forthe period of record yields the following data: 1) meanTABLE 2. Statistical summary of 2 h winter andsummer temperatures (-F).Arithmetic meanStandard deviation Winter data [19 December 1974 to 25 January 1975 (372 cases)]1 2 3 4"Beach" "Institute" "Elgie" "Welder'57.41 58.14 56.55 56.737.68 7.96 10.82 11.65Arithmetic meanStandard deviationSummer data [-6 June to 10 July 1975 (316 eases) ~83.30 82.69 82.00 81.99 2.43 3.30 5.01 6.42JULY1977 KENNETH H. JEHN 913temperature of the warmest month (August), 83.9-F;2) mean temperature of the coldest month (January),63.5-F; 3) annual range of temperature, 20.4-F. Thecalculation of continentality index (Conrad) givesk= 17.6%, a not unreasonable overwater value for thecoastal waters approximately 80 mi off the TexasCoast. A similar calculation for the Galveston area(Area 28, at 28.3-N, 93.2-W) gives k= 20.0%. This isin general agreement with the previous finding of largervalues of continentality along the upper Texas coast,compared to the lower coast. Moreover, these over watervalues of continentality seem to fit reasonably with thebarrier island value of 20% estimated for the coastalbend near Port Aransas.3. Summary and conclusions A preliminary, distribution of continentality index bythe Conrad definition has been exhibited for the Texascoastal zone, based on climatological data for the period1941-70. Additionally, detailed field measurements ofsurface air temperature from the Gulf beach at PortAransas inland to near Sinton in a summer and a winterperiod have been used to demonstrate relatively largehorizontal gradients of continentality index in the Texascoastal bend. Finally, estimates of continentality indexover water have been made from available surface airtemperatures made on ships over the period 1952-71.Internal consistency is exhibited by these various continentality values. Continentality by any definition is but one of manyparameters used to characterize the climate of a givenarea. It is clear from this study that detailed distributions of climatic variables will be needed to characterizeclimatic resources of the Texas coastal zone, especiallyin the region closest to the coastline, including bays andestuaries and the adjacent Gulf, where very few dataexist. Resource catalogs of the Texas coastal zone suchas those of Fisher et al. (1972) and Fruh et al. (1972,1973) are deficient in their listing of climatic resourcesas a consequence of dependence on existing climatological stations. Data needs can be met in part by studiessuch as the one described here. A better long-termsolution would be the establishment of additionalclimatological stations close to the coast. A good starting point would be to establish a standard climatologicalreporting station at the Port Aransas Laboratory of theUT Marine Science Institute, and the simultaneousplacement of an automatic weather station (AMOS) onone of the oil or gas platforms just offshore. The datafrom these two stations would serve present operationalneeds of the Corpus Christi forecast district, and wouldbe the nucleus of additional studies of the transitionalclimates of the Texas coastal zone. Acknowledgments. The author has received assistancefrom many persons and organizations during this work.Chief among the organizations are the UT College ofEngineering Bureau of Engineering Research, whichprovided a grant for part of the research; and the UTMarine Science Laboratory at Port Aransas, which provided laboratory and storage space for equipment, andvehicles for transportation in the field. My wife Peggy has been a capable research assistantin much of the field work. J. C. Evans assisted with theinitial study, psychrometer transects by automobile'from Port Aransas to ~20 mi inland, 15-16 August1974. Larry Powers helped with chart changing and inother ways. Dr. L. Drawe maintained records at theWelder Station. Leif B. Svensson assisted with theliterature search. W. Randall Poteet made most of thecontinentality calculations. Sue Sweeney typed themanuscript. REFERENCESBailey, H. P., 1968: Hourly temperatures and annual range' Eclectic Climatology, A. Court. Ed., Oregon State Univeristy Press, Corvallis, 184 pp.Berg, H., 1940: Die Kontinentalit~tt Europas und Ihre Orderung 1928/37 gegen 1888/97. Ann. Hydrogr. Berlin, 124..Brunt, D., 1924: Climatic continentality and oceanity. Geogr. J. London, 44, 43-56.Calef, W., 1950: Interdiurnal variability of temperature extremes in the United States. Bull. Amer. Meteor. Soc., 31, 300-302.Conrad, V., 1946: Usual formulas of continentality and their limits of validity. Trans. Amer. Geophys. Union, 27, 663-664.Court, A., 1951: Temperature frequencies in the United States.J. Meteor., 8, 367-380. Craddock, J. M., 1964: Intemnnual variability of monthly mean air temperatures over the Northern Hemisphere. Sci. Pap., Meteor. Off. London, No. 20.DeBrichambaut, G. P., and C. C. Wall~n, 1963: A study of agroclimatology in semi-arid and arid zones of the Near East. WMO Tech. Note 56, 19-22.Dinies, E., 1932: LuftkiSrper-Klimatologie. Arch. Dtsch. Seewarte, 50, No. 6. D'Ooge, C. L., 1955: Continentality in the western United States. Bull. A~ner. Meteor. Soc., 36, 175-177. Fisher, W. L., J. H. McGowen, L. F. Brown, Jr., and C. G. Groat, 1972: Environmental Geologic Atlas of the Texas Coastal Zone. Bureau of Economic Geology, The University of Texas at Austin, 7 volumes. Fobes, C. B., 1954: Continentality in New England. Bull. Amer. Meteor. Sot., 35, 197. Fruh, E. G., et al., 1972: The management of bay and estuarine systems, Phase I: A conceptual report. Division of Natural Resources and Environment, The University of Texas at Austin. , 1973: ;rhe management of bay and estuarine systems, Phase II: A conceptual report. Division of Natural Resources and Environment, The University of Texas at Austin. Gorczynski, Wl., 1920: Sur le calcul du degr~ du continentalisme et son application dans la climatologie. Geogr. Annaler, Stockholm, 324-333. Hann, J., 1908: Handbook of ~limatology. Stuttgart, J. Engelhorn, 215-216. Hela, I., 1953: Regional distribution of the continentality in the climate of the oceans. Geophysica, 4, 41-47. Jehn, K. H., 1974: The role of the atmospheric sciences in the Texas coastal zone. Rep. No. 40, Atmos. Sci. Group, The University of Texas at Austin, 70 pp. Johansson, O. V., 1931: Die Hauptcharacteristik des jahrlichen Temperaturganges. Gerlands Beitr. G-ophys., 33, 406-428.914 MONTHLY WEATHER REVIEW VOLUME105Kerner, F., 1905: Thermoisodromen, Versuch einer kartographis chen Darstellung des jahrlichen Ganges der Lufttemperatur. K. K. Geogr. Gesellsch. Wien, 6, No. 3.Kopec, R. J., 1965: Continentality around the Great Lakes. Bull. Amer. Meteor. Soc., 46, 54-57.Leighly, J. B., 1938: The extremes of the annual temperature march with particular reference to California. Univ. Calif. Publ. Geogr., 6, 191-234. , 1941: Effects of the Great Lakes on the annual march of temperature in their vicinity. Pap. Mich. Acad. Sci. Arts Lett., 27, 414.MacKay, D. K., and F. A. Cook, 1963: A preliminary map ofcontinentality for Canada. Geogr. Bull. Canada, 20, 76-81.Polowchak, V. M., and H. A. Panofsky, 1968: The spectrum of daily temperatures as a climatic indicator. Mon. Wea. Rev, 96, 596-600.Prescott, J. A., and J. A. Collins, 1951: The lag of temperature behind solar radiation. Quart. J. Roy. Meteor. Soc., 71, 121.Raunio, N., 1948: The effect of local factors on the meteorological observations at T6rshavn. Geophysica, 3, 173-179.Spitaler, R., 1922: Klimatische Kontinentalit~tt und Ozeanit~tt. Petermann' s Mitteilungen.Trewartha, G. T., 1961: The Earth's Problem Climates. University of Wisconsin Press, 254 pp.U. S. Naval Weather Service Command, 1975: Summary of Synoptic Meteorological Observations, North American Coastal Areas (rev), Atlantic and Gulf Coasts, Vol. 4, pp. 554-632 (Table 15).