Application of Hydrology to Irrigation, Drainage and Soil Conservation in Japan (Part I)
By KINJIRO SHIBUYA
Water Management Division, National Research Institute of Agricultural Engineering
drought is liable to occur in the year which Introduction has scanty rainfall. Irrigation planning is usually made against such a droughty year Application of hydrology to agriculture as occurring several times in several-ten is found in irrigation, drainage and soil con years, so that irrigation water can meet the servation. requirements for most years. Agriculture, which is carried on in natural The average recurrence interval at which environments for the most part, varies in its more serious drought than the planned one water use from country to country. The way will occur is officially set to be ten years for how the hydrology is applied to agriculture, major irrigation projects in Japan. Actually, therefore, will differ according to the natural the basic year for planning is determined by environments of each country. selecting from the past years one of such The hydrological techniques which are in years which had scanty effective rainfall, ex practical use in Japan will be presented briefly tended dry spell, and/or great damage which in this paper, with a hope that some of them would occur once in ten years on the average. may serve as useful information to other (2) Effective rainfall countries. Effective rainfall for paddy fields was for merly estimated to be 80% of daily rainfall between 5 and 80 mm in an irrigation period. Application of hydrology to This estimate probably includes the utilization irrigation of rainfall detained in dual-purpose canals. The recent separation of irrigation and Hydrology is applied to the estimation of drainage canals reduces the utilization of the effective rainfall, evapotranspiration, stream storage, so that the amount of effective rain flow, and inflow into a reservoir. fall seems to become a little smaller than be fore. Some field observations indicate that 1) Bci.~ic yea1· for planning daily rainfall between 5 and 30 mm corres (1) Recurrence interval of the basic year ponds approximately to the effective rainfall, for planning but further researches have to be made for Water requirements of paddy fields in Ja reasonable estimation in various situations. pan are 1000- 1500 mm/yr or 1000- 15000 m3/ Effective rainfall for upland fields has to 10 a/yr on the average. Of the total water be estimated by making water budget. The requirements, 200-400 mm/yr or 200-400 m3/ upper limit of available moisture in soil cor 10 a/yr is usually supplied from rainfall. responds to the amount of moisture held at This portion of rainfall is called effective field capacity, and the lower limit of available rainfall. In a region, where streamflow, from moisture is the amount of moisture retained which irrigation water is diverted, is posi at wilting point. Actually, the total readily tively correlated to rainfall, a serious available moisture (TRAM) should be adopted 32
Table 1. Evapotranspiration ratio Er/Ep in Japan ( ¢>20 cm evaporimeter) Month Surface condition Jan. Feb. Mar. Apr. May June July Aug. Sept. Oct. Nov. Dec. Paddy field 0.5 0.5 0.6 0.6 o. 7 1. 0 I. 2 I. 3 l. 1 o. 7 0. 7 0.6 Upland field 0.5 0.5 0.6 0.6 0.7 0. 7 0.8 0.8 0.8 o. 7 0. 7 0.6
as the upper limit of available moisture. The year which had such drought damage The estimation of 10-year·» effective rainfall as would occur once in ten years on the aver is usually made by fitting the effective rain age should be adopted as "a proposed basic fall data to the normal distribution and by year for planning". reading off the value co1Tesponding to 10% of cumulative frequency. The year which had 2) Evapotransviration effective rainfall approximate to the 10-year Evapotranspiration occupies a large part in value should be adopted as "a proposed basic consumptive use of water on pady fields as yea1' for planning". well as on upland fields. (3) Successive rainless days Inflow into a reservoir also is normally de A drought may occur owing not only to pendent on water losses which consist mainly scanty effective rainfall but also to an ex of evapotranspiration from the watershed. tended dry spell. Thus, the estimation of evapotranspiration The statistical analysis of the number of under various conditions is indispensable to successive rainless days is usually made by the inigation planning. fitting the data to lognormal or extremal The methods of its estimation are classified distribution. into (1) evapotranspiration 1·atio, (2) water The year which had the number of succes balance, and (3) heat balance. sive rainless days approximately correspond (1) Evapotranspiration ratio ing to the 10-year value should be adopted as Evapotranspiration ratio is a ratio of "a proposed basic year for planning". evapotranspiration E,. to evaporation from pan The method of statistical analysis in this evaporimeter EP, as shown in equation (2) and relation will be explained later. Table 1. (4) Drought damage E1· Drought damage data are usually collected f = Ep ...... (2) for several-ten years. The damage may be expressed either by income deficit or by yield Ep is usually measured with 20-cm pan reduction of main crops. evaporimeter. The average recunence interval assigned to The ratio, however, may change according a certain value may be computed using equa to climate, vegetation, and cultivation. tion (1). (2) Water balance N+l Evapotranspiration can be estimated by T,=- .- ...... (1) t solving water-balance equation after taking field observations of each of its terms except where T1 : the average recurrence interval evapotranspixation itself. for i-th largest damage; N: number of the The watershed on which evapotranspiration damage data. is to be estimated must meet such conditions * In this context and the like, " IO-year" means that it has no deep percolation, no leakage, average recurrence interval of 10 years. measurable rrround\·1ater flow, and negligible 33
Table 2. Specific discharge of droughty flow in Japan Specific discharge Specific discharge Region in m3/ sec/ 100 km2 in mm/ day
Northern coastal area of Japan more than 2. 7 more than 2. 3 Central mountainous area about 2. 7 about 2.3 Coastal area on Pacific Ocean about l. 8 about 1. 6 Mountainous area of south Kyushu more than 2. 7 more than 2. 3 Central area of Kyushu J. 2- 1. 4 l. 0-1. 2 Seto-inland Sea area about 0.9 about 0.8 Shikoku Island 0.9- 1.8 o. 8- 1. 6 Southern part of Kinki district 1.0- 1.8 0. 9- 1. 6 Southern costal area on Japan Sea about 1. 8 about 1. 6 change in watershed storage. droughty years is very important in irriga This method is suitable for the estimation tion planning. of long-term areal evapotranspiration. (1) Estimation of droughty discharge of (3) Heat balance streamflow Evapotranspiration can be estimated using The droughty discharge, if there are ade heat balance equation, as shown below. quate data available, can be estimated using lognormal or weibull distribution. If there R-G ( r Ti-Tz ) E =- L~ 1- A -T--T- ...... (3) are few data available, it has to be estimated u+r 101- w2 using some hydrological techniques which will where E: evapotranspiration in g/cm2/min; be explained later. R : total net radiation flux in cal/cm2/min; The droughty discharge typical of Japan is G: heat flux into the ground in cal/cm2/min; shown in Table 2. L: latent heat of vaporization of water in (2) Estimation of inflow into a reservoir cal/g; y: psychrometric constant, f>.: slope of The capacity of a reservoir must be deter vapor pressure curve at the average of T,.1 mined so as to meet maximum cumulative and T .• 2 ; T,.,, T"2 : wet bulb temperature at deficit of irrigation water in the basic year the two heights; T 1 , 1\: dry bulb temperature for planning. If there are adequate data on at the two heights. inflow, the computation of 1·eservoir capacity This method has recently been put to practi can be performed without difficulties, but if cal use owing to the development of a new type there are few data available, inflow must be of actinometer with polyethelene film cover. estimated using hydrological techniques such The polyethelene film is almost entirely trans as tank model. Moreover, if there are no data parent for both long and short wave radiation. on inflow except some data on rainfall at the This property makes it possible to measure site in question, rainfall for longer term must total net radiation flux easily and accurately. first be estimated from the rainfall data at This method is hoped to be applied to nearby sites by means of equation (4). further researches on evapotranspiration in - Sv - various situations. Y = Y+s;r(X-X) +ue ...... (4)
3) Estimc,tion of water resources for irri where Y: estimate of daily rainfall for the gation site in question; Y: mean value of y; x : daily Water requirements of farm lands must be rainfall at nearby site; x: mean value of x, r : met by effective rainfall and irrigation water correlation coefficient between x and y; S,:
taken from reservoirs and streamflow. Hence, standard deviation of x; S1 : standard devia the estimation of those water resources in tion of 11; u: mean residual; e: random num- 34 JARQ Vol. 14, No. .l, 1980
ber generated from N (0, 1) ·>:-_ (iii) In alluvial or delluvial fans where G2- G~ is commonly negative, i.e., the seeping 4) Water balance analysis out of groundwater takes place, Dc D 2 is (1) Water balance equation smaller than E. Water balance equation for a delineated (5) Estimation of return flow area and period is as follows: Outflow of surface water D 2 may be sup P= E+(D2-D1) + (G2-G1) +L1S} posed to be return flow, if inflow of surface (5) L1S= L1H· P. + L1M + L1 W, ... water D 1 is at reasonable rate. But the re where P: precipitation; E: evapotranspira turn flow on which downstream paddy fields
tion; Du D 2 : inflow and outflow of surface can rely has to be a steady flow after such a
water; GJ, G2 : inflow and outflow of ground long dry spell as anticipated in planning. A water; AS: change in storage; AH: change in careful study must be made of steadiness of groundwater level; A.M: change in soil mois return flow using water balance equation. ture in unsaturated zone; Aw.: change in (6) Change in storage storage of surface water; P,: effective Manipulation of storage of surface water is porosity of aquifer. an important part of water management for (2) Estimation of evapotranspiration by paddy fields. Thfa must, however, be done water balance taking into account the influence upon sur As mentioned in (2) of 2, evapotranspira face flow, groundwater flow, and storage in tion from a watershed can be estimated using soil and groundwater. equation (5) under the condition of 6s =:: 0 , • 5) S erial storage model or tank model and with observations of P, Di, D 2 G,, and G2 This condition will often be met by headwater This model which was proposed by M. Suga area with outcropped measuring site. wara is frequently used today for water re (3) Estimation of groundwater flow sources engineering. In equation (5), for a period with As :::,: 0,
field observation of D 1 , D~ and P together r(t) with estimation of E may reveal groundwater i h flow G2-G,. G2-G1, if positive, means the recharge of groundwater from surface water in the area. /\2 ... ( 4) Some results of water balance analysis h(t) [ in Japan (i) In paddy field areas on low-lying allu "· I ,.r ~· vial plains, G2- G, is small owing to small hy q(t) q draulic gradient of groundwater, and the value t Ao of D 1- D2 in equation (5) approximates to E i(t) regardless the size of the area. q(t) = q,(t)+qit) h(t) : Storage (mm) ii) In paddy field areas, where almost all q1(t)=(h(t)- H1) · l 1 Ao: Coeff. of seepage hole water percolated into ground flows out of the qi{t) =( h(t)-H2) • A2 J,, l2: Coef. of discharge hole - i(t) = h(t) · l 0 r(t) : Input (mm) area as groundwater flow, the value of D 1 D2 in equation (5) approximates to average i(t) : Output seepage (mm) r(t)-q(t)-i(t) =9.h. q(t): Output discharge (mm) water requirements of paddy fields over the dt area, but the larger the area, the nearer to Fig. I. Operation of the tank model E it becomes. The operation rules of this model can be * Normal distribution with mean= O, and variance seen in Fig. 1. = 1. The working procedure for calculation using 35 this model will be explained briefly as follows. (1) Zoning of watershed 20-30 o.oa-o.,, Since a long-term rainfall may increase ac t 0-10 l 0.03-0.20 cording to the altitude, the zoning of a water j shed should be carried out into 4 zones or 0.15-0.35 so, according to difference in altitude within the watershed. Then the correction factors for each zone should be assumed using rele ~w LJo,-oos vant data. (2) For seasonally-snow-packed area, snow 0.06-0.13 melt calculation should follow. (3) The next step is the estimation of ,-.,,LJo.oo,-om, areal evapotranspiration, which may be made using reduction factor of 0.5-0.8 together 0.006-0.08 with evaporations from