204 MONTHLY WEATHER REV1E.W. APRIL,1921 such a drainage area. It amounts to a run-off depth of Obsertrers both at Sand Lake and Taborton state that onl .0.16 inch per hour. There are, however, two lakes 29 years ago, in Au ust, roads were washed out and in &is basin, and the amount of stream channel storage is streams were slightly &her than on the recent occasion. relatively large, so that from the same rainfall a run-off Probably the intense rainfall of Au ust 10 covered only rate much larger would natural1 be expected from the the higher easterly portions of the %and Lake Drainage land area tributary to Bowman Jond. Basin. (See fig. 3.) CORRELATION OF WIND VELOCITY AND CONVECTIVE RAINS AT HOUSTON, TEX. By I. R. TANNEHILL,Obaerver. [Weather Bureau, Houston, Tex., Apr. 16,1921.1 SYNOPSIS mined. Ita eastrwest length is approximately 700 miles, or from the Atlantic in the neighborhood of Jackmnville. Fla.. to abput Houston, In the Vicinity of Houston, Tex., convective rain6 are frequent Tex. The pressure conditions &moCiStd with these ra~mare about during the Bummer monthe. These rains result from expaneion and as follows: consequent cooling of air rising in a nearly vertical column. The air A high, With pressure 30.15 to 30.20 inchea, overlim the southern can not hein a continuous convection column if the wind is of more portion of the middle Atlantic, with an exteasion over the Florida than modem& vel@y nor can it rise through or even well into a peninsula, in which the preseure is 30.08 to 30.10 inches. Preeaiue StdUm Of Et- winds. diminiahea in a weaterly direction to a region of indifferent gradients From a @ud of wind movement and eminfallat Houston. Tex., it is over mutheastern Louisiana. The gradients are for gentle miithast a parent that &ere& a stron relqtionslllp between these two elements. winds along the coast and over the narrow fringe of the interior. &e lighter the dcewin% movement the greater is the probability of local convective rain^ in summer. In the foregoing quotation it is pointed out that the In the vicinity of Houston there is suficient mohture on nearly all pressure distnbution such as to give rise to gentle days in aummer: the neceeaa condition for the formation of convec- is tive rain^ & local heating in rxatively quiet air. southeast winds. The temperature of the air does not differ materially The majority of the rains which occur at Houston, Tex., from one summer day to another in .this section. On in summer are of a type eculiar to the territory bordering nearly all days there is Fn abl)ndanced moisture. The the Gulf of Mexico. fhe southeasterly monsoon is a important consideration seems to be the velocity of the source of abundant moisture on the coast. Toward the southeast wind. When the wind is strong,,rain is pre- interior the rainfall dminishes. In other portions of the vented by mixing, and the moisture is carried farther State the precipitation of this moisture may be due to int.0 the interior. cyclonic action or to forced ascent. In the coast section The records at Houston, Tex., covering a period of cyclonic disturbances in summer are limited mostly to 11 years, 1910 to 1920, inclusive, were examined in an rather infrequent tro ical storms and since that section effort to determine any relationship that exists between is practically flat, littP e of the rainfall can he ascribed to rainfall and wind. During this period the anemometer the forced ascent of air currents. Nearly all of the rains and the rain gage have not been moved. which do occur in the vicinity of Houston are undoubtedly convective in origin. TABLEl.--,41remge wind velocity, mika pw hour, and the number of The strength of this southeasterly wind is variable. rainy days, 0.01 inch or mare, for each of the mn#a Jum, July, and With the passage of high-pressure areas to the northward August. this monsoon becomes feeble and sometimes t.here occur distinct land and sea breezes. On a typical sumnier day, June. July. Alwt. however, the southeasterly wind continues throughout t.he I I I 24 hours, becomin stronger after midday. The forma- Avera Number tion of cumulus couds IS an almost daily occurrence. Year. winP T move- or days Yet day after day asses in summer with no minfall at ment withO.01, Or- znT (miles (miles Houston if the sout easterly wind is strong. ralnfan' per hour) On some days theR convective column builds upward to v-) great heights in uiet air. On other days the tops of the 1910 ...... 7.4 9 7.4 11 7.1 6 cumuli appear to e torn away and to dissolve on enterin 1011 ...... 7.1 9 6.7 13 6.8 9 % 1912 ...... 85 9 6.3 12 6.9 11 a stratum with a higher velocity of .Tdmovement. fi 1913 ...... ai 9 7.5 5 6.3 10 1914 ...... 6.7 7 6.7 5 6.7 18 is obvious that a strong wind, by mixing unsa.turated air 1915~~ ...... 82 a 7.9 s a3 17 with the air in the column, tends to prevent further growth 1916 83 7 5.9 13 6.6 14 ,1917 a7 3 7.6 7 6.8 4 of the cloud mass. Quiescence of the upper air is there- 1918 6.9 5 6.7 6 6.6 9 1R19 7.4 13 LO 12 5.8 11 fore an aid if it is not on some occasions essential to the 19m 6.7 10 6.1 12 5.5 17 formation of convectional columns to great altitudes. -- , quiescence of the surface air favors the convections, since it prmits, as strong winds For the values in Table I, the coefficient of correlation I do not, appreciable local inequalities of temperature. has been corn uted by the method of least squares. The From observation it therefore appears that the prevail- equation uselwas ing southeasterly wind at Houston is productive of ahun- dant rainfall when conditions are favorable for convection and that the important consideration is tho st,rcngth of thi? southeasterly current. for the correlation coefficient aiid Concerning convective rains, Prof. A. J. Henry, in Weather Forecasting in the United States, says: 1-9 E, = 0.674 - There are two districts in which convective rsinS occur during the dri warm 8e88on. The first of the districts is alo the Gulf coast, inclucl- - in the Florida eninaula, and extend' bax into the interior rob 1 See: The Effect of Weather upon the Yield of Corn, Mo. WmTnLI REV. Feb., 1914, ebfy not more $n 50 milee, the exacxoder... not being... 88 yet fete; 42 7e-FI; also Elementary Now on L+st Squama, e.tc., ibid., at., 1914, d:551-588.

Unauthenticated | Downloaded 09/30/21 12:06 PM UTC .4PRlL, 1021. MONTHLY WEATHER REVIEW. 206 for the probable error of the coefficient, where z is the From the values in Table 21 the coefficient of correla- departure of wind velocity from its mean and y is the tion is found to be - 0.81 *0.07. This is positive assur- de arture of,number of ram days from ita mean. ance of a relationship, although the higher value is to be gram these data the coe& cient. is found to have the espected, on the assumption that there is such a relation, ualue - 0.42 fO.lO. This is a fairly high value. It will because of the longer periocl for which the avera ea were be noted that one'unexpectedly large wind movement t,aken, bein annual instead of monthly as in Ta% le 1. occurred in connection mth a lave number of rain days These vauesY of the correlation coefficient a bear out, in Au st, 1915. This value is sRown.at "A" in gig. 1. the contention that the rains are more frequent in periods This Fgh wind movement was the result of the hurri- when the wind is light and also indicate that the season cane during that month. If 'that month's record is of convective rains sometimes extends into other months omitted the value of the coefficient of correlation is found of the year. to be - 0.56 f 0.08, in which the coefficient is seven times the probable error. TABLE3.-Awraqe daily ridnd moamnt, milea per how, of July, 1910- Because the influence of ext.ratropica1cyclones has not 1920, inchive, .fm days with rainfall of 0.01 inch or more and for aayS yet ceased in June, and because of the frequency of with rainfall of 1 inch or nwrc. tropical storms in this vicinity in August, the coefficient $3WinB movement. For all dap...... 6.7 miles. For days with rainfall, 0.01 inch or more...... 6.0 milea. For days with rainfall, 1.00 inch or more ...... 5.4 mil-.

The averages in Table 3 prove beyond question that the wind movement is less ra id on rainy days and that the variation in the wind veP ocity is inversely propor- tional, to a certain estent, to the amount of rain. The foregoing discussion is based upon surface wind movement. The wind in the lower levels of the atmos- here usually changes rapid1 with altitude. The sur- face wind is not a reliable in ex'to conditions aloft. If the above relation is true andB it a pears to have been established, it remains to study tg e relation between the wind aloft and the rainfall.J With'increasing height above ground the wind be- comes more and more nearly constant. The wind st some distance above the surface is therefore a more Number Of Days m Month wjfh O.O/ reliable indication of the eneral air movementa over w mare of Pr8Ciphfion. the region in question. ALowledge of the wind aloft is therefore important. Fio. l.--Relation between number of dny da and wind movement in June, July, and August, 191C1920, incl&, at Houston, Tex. of correlation has been computed for the month of July, CONCLUSION. for which the data are found in Table 1. The value for Jul is -0.65f0.12, in which case the coefficient is In the vicinity of Houston, Tex., where there is in higi er and the probable error larger, which is to be summer abundant moisture and the temperature is expected. ersistently high, strong winds hinder the formation of TABLE2..-Av?age. wind velocity, miles per how. for. each gear. 1910. tn kcal convective systems while any stagnant condition 1920, melmve, and the annual number rainy day8 for the 8amc of the lower air is likely to roduce fre uent rains. A +- consideration of the gener2 pressure 1istribution to- gether with a study of wind movement aloft should Year. Iz3wind. days. enable the forecaster to predict convective rains with greater accuracy. Mi.lHr. Mi.f Hr. -- 1910 ...... 8.4 93 1916 ...... 8.4 87 9 In the Mo. WEATHERREV. Feb. 1914 42: 79, Prof. J. Warren Smlth says: 1911...... 8.n 115 1917 ...... 8.6 ti9 "It probabl is safest to ashme that &ere may be some relation if the correlation 1912 ...... 8.3 100 1918 ...... 11.2 93 roRfRrient is dree tim& the probable error and that the relation is established beyond 1913 ...... A.3 108 1919 ...... 7.7 126 qurstion if it is more than sir times the probable error." 1914 ...... 8.1 112 19m ...... 8.2 113 a The vertical temperature graliznt ii important but little is known co1mrning tem- . 1915...... 8.6 89 peratws aloft at Houston during the convectlve rain ~aso11.

Unauthenticated | Downloaded 09/30/21 12:06 PM UTC