Taiwan Today

Everybody knows the importance of water in his life, as he drinks and washes everyday, sprinkles his lawns and waters his gardens every season, and enjoys boating, fishing and swimming during weekends; but very few people are quite aware of what scarcity of water would mean to a Taiwanese. Indeed, one can say that the people in Taiwan live entirely upon water. Any deficiency in rainfall like last year's would mean dry rice fields, shut-down factories, and blacked-out night hours; and it would also mean high prices of food, lost wages, and extra expenses for candles or oil lamps. Taiwan, however, is far from being an arid or semi-arid region. With an average annual rainfall of nearly 100 inches, Taiwan is certainly ranked high among regions classified as wet. And there are records of rainfall in Taiwan which are not far below the world's highest. For instance, a 24-hour rainfall of 45 inches was recorded at Kuwarls (July 6-7, 1931) while the world's highest was the 46 inches recorded at Baguio, P. I. in 1911. Therefore, scarcity of water in one season is often compensated by an excess in another. This is exactly the case in many parts of Taiwan, which is indicative of the climatic conditions of the monsoon area in South Asia. The records of three stations are given below as examples:

These records clearly show how a wet country experiences an extremely dry period each year.

On the other hand, the northern part of the island has a much more uniform distribution of rainfall, as may be seen from the records of the other three stations:

Unfortunately, this northern tip comprises only about 13% of the total area of the island, and one can say that in general Taiwan has a climate of alternatively wet and dry seasons.

Speaking of its rivers, Taiwan has the worst of all. Running down from mountains 10,000 to 13,000 feet above sea level, the longest of its 152 drainage chancels is only 105 miles, with an average slope of 1 in 59 or 89.5 feet per mile! The channel having the gentlest slope drops 8.5 feet per mile. One can see vary well from these figures that all the rivers in Taiwan are torrential in character, draining quickly whenever there is a rain, without any detention or reservation. In spite of thickly forested areas, observed coefficient of run-off or percentage of rainfall that flows down the river as surface water has been as high as 96.65% and very rarely below 70%. Consequently, rivers swell high during a storm and shrink nearly to gutters after a short period of sunny days. Of the 19 most prominent rivers, the minimum measured flow is only about 3.5 c.f.s. (cubic feet per second), while the maximum reaches 780,000 (calculated) c.f.s., or more than one-fourth of the flood discharge in the Mississippi River. Or, expressed on unit area basis, the maximum flow of the large rivers in Taiwan amounts to 264 to 755 c.f.s. per square mile, as compared with 27 to 324 c.f.s. per square mile of American rivers of the same size. The ratio of maximum to minimum flow ranges from 440 to 30,000.

Another unfavorable condition for the development and conservation of water resources in Taiwan is the geological structure of the island. Composed mostly of recent formations of rocks and subjected to frequent disturbances by earthquakes, practically all the slopes of high mountains may slide down to block the water courses and add a vast amount of bed load to the stream flow. Two most serious slides occurred successively on December 18, 1941, and August 10, 1942, at an abrupt turn of the creek draining the Ali Mountains. These slides blocked up the stream to form a reservoir about 550 ft. deep with a dam of earth and boulders estimated at about 130 million cubic yards. This dam had been preserved with a view to utilizing the God-given reservoir until May 1951, when a not too great storm carried a large quantity of earth down the stream in the already eroded outlet channel, thereby emptying the reservoir within a few hours and causing the loss of more than a hundred lives. Small-scale slides occur everywhere along the rivers, raising the riverbeds, obstructing bridge openings and burying power plant structures. Measures of soil conservation are often ineffective owing to unstable slopes loosened by earthquakes, or too expensive owing to the extensive area involved in each case.

Notwithstanding the hindrances and difficulties mentioned above, the water resources of Taiwan are still much richer than that in many other parts of the world, and their conservation and development must be wisely included in any general program far promoting the well-being of the people in this island. Early in its history, farmers from the mainland of China settled in this island and brought with them the technique of building canals and irrigating rice. In two and a half centuries (1655-1895), the population increased rapidly and the area of irrigated land increased to nearly half a million acres, or 56% of the total cultivated area. After the Japanese occupation, the development of this island was much accelerated. The census of 1943 showed the population to have increased from about three million to more than six and a half million, and the cultivated area from one million to two and one-tenth million acres. The irrigated area totaled about one and one-third million acres, or nearly 62 % of the cultivated area. This percentage had been brought up to 70% in 1952 with an increase of about a quarter of a million acres.

This is certainly a remarkable achievement as compared with many other countries. For instance, India, a nation very famous for her extensive irrigation systems, has only about 14.5 acres of irrigated land per 100 persons, while in Taiwan it is about 20.3 acres in 1952. Only Pakistan surpasses Taiwan in per capita irrigated area - 28.5 acres per 100 persons. But Taiwan produces a much greater crop per unit area than both India and Pakistan because of the intensive cultivation and the use of fertilizers. Its average production of 2,200 pounds of rough rice per acre per crop (1934-38) was about 67-88% higher than that of Pakistan and India (1,315 and 1,168 pounds respectively). The actual per capita production of paddy in these three countries in 1949-50 was: Taiwan 415 lbs, Pakistan 366 lbs, and India 205 lbs. If 500 pounds of paddy is taken as the minimum average consumption per capita per annum, it is no wonder that in a region with less than one half the density of population of Taiwan like India and Pakistan (258 per square mile), serious famines often break out while Taiwan has surplus rice to export to other countries.

In another respect, namely, the annual consumption of electricity per capita, Taiwan also surpasses all her neighbors but Japan. In 1950, 1.28 billion KWH of electricity was generated in Taiwan, or 168 KWH per capita. In the same period, it was 120 in Malaya, 24 in the Philippines, 15 in India, 2 in Pakistan, and 446 in Japan. The power in Taiwan came mainly from hydroelectric plants; only 9% was generated by thermal plants in that year. The development of hydroelectric power in Taiwan had its beginning in 1905, when a small plant named Kwei-Shan generating an installed power of 600 KW was completed. Gradual expansion of the industry and the completion of a few large plants like Sun-Moon Lake plants No.1 and No.2 brought the total installed capacity to 267,000 KW in 1945, but war and floods destroyed most of them, and only about 40,000 KW could be generated at the end of World War II. By 1951, however, restoration and reconstruction had practically recovered the full capacity, which will be further increased after the completion of the Tienlun plant.

The development of public water supply has been going on hand in hand with the development of water power. Large and small, there were 140 waterworks in the island in the year 1951, serving a population of 2,160,000, or 29% of the total. The present per capita consumption averages 175 liters per day, the total consumption being 153 c.f.s.

Altogether Taiwan at present uses about 45,000 c.f.s. during the full growing season for both irrigation and domestic water consumption, while the average flow of the "whole island is about 100,000 cubic feet per second. The maximum capacity of the present waterpower plants requires only about 10,000 c.f.s. In fact, according to the estimate of the provincial Water Conservancy Bureau, only 19 per cent of the water resources is now utilized. In other words, 81% of this form of natural wealth is still being wasted. Moreover, that portion already developed is by no means utilized efficiently, as shown by the very low duty of water in record. (The duty of water in irrigation means the size of the area served by a unit discharge like c.f.s.) In India, the duty for rice fields is 80 acres per c.f.s., while in Taiwan it is less than 40 acres. The wasting of water, especially during storm flow, also means the loss of soils and soil fertility, and the raising of riverbeds which is accompanied by increasing of flood hazards and losses. Therefore, the conservation of water is much more important than anything else in this region. This may be done in two ways, viz., the conservation of flow and the diminution of losses and wastes. Regarding the former, we can enumerate soil conservation, reservoir construction, and ground water recharge. Regarding the latter, improvement both of methods of irrigation and of canals and canal structures is of the utmost importance.

II

The storage of floodwater for utilization in the dry season had been practiced jointly with the building of irrigation canals. These farm-ponds, small and scattered, constitute au important source for watering many acres of rice fields not included in the irrigation systems. One big system of impounding water had been constructed along with the Taoyuan canal which was completed in 1924, irrigating 52,500 acres of plateau land south of the town of Taoyuan. There are now 241 ponds storing a total of 39,502 acre-feet of water that can give a flow of 275 c.f.s. during irrigation seasons. The water is supplied mainly by the irrigation canal in winter time. Ten reservoirs had been built between 1930 and 1943, with a storage capacity ranging from 472 to 132,500 acre-feet. The largest is the Coral Lake, created by an earth dam 184 feet high across a small branch of Tseng-wen Creek, and supplied from the main river by a tunnel. This reservoir irrigates a great portion of land under the system now called the Chia-Nan Canal. The second largest reservoir is the Sun-Moon Lake for the development of power. It makes use of two natural depressions and increases the storage capacity to 118,000 acre-feet by two comparatively low earth dams in the mountainous reaches of the largest river in this island, the Cho-Shui Creek. Water is diverted from the upper branch by a long tunnel, and a two-stage power development generates a total of 143,500 KW. All these reservoirs are for the single purpose of water utilization, and store only a little portion of the floodwater as their inflow is all limited by tunnels.

The completion of the Akungtien Reservoir in 1952 marks a new epoch in the history of water conservation in Taiwan because it is a so-called multiple-purpose development. Though small in capacity - only a little more than one-fourth of that of the Coral Lake - it will serve to control the flood water from the Creek bearing its name, thereby saving the town of Kangshan in Kaohsiung District from annual inundation and giving it an excellent domestic water supply. It will also irrigate an area of about 5,300 acres. In addition to Akungtien, 16 multiple-purpose reservoir sites have been proposed, with a total capacity of 2,500,000 acre-feet, together with 30 single-purpose ones. One which may be mentioned here is the Shihmen or Stone Gate Reservoir on the Ta-Ko-Kien Creek, a branch of the Tamsui River. This reservoir and dam site has been investigated and surveyed during the last few years, and the building of a dam about 480 feet in height will create a full storage capacity of 470,000 acre-feet, of which 80,000 is for dead storage, and the rest for flood control, irrigation, and power development. The estimated cost of this dam and reservoir project would be from US$40,000,000 to US$55,000,000. The benefits to be derived therefrom would be: (a) an additional irrigated area of 110,000 acres for double crops of rice, (b) generation of electric power estimated at 27,000 to 60,000 KW, and (c) reduction of flood height.

Reservoir development in Taiwan is, however, a costly undertaking. Let us take the Shihmen Reservoir for example; every acre-foot of storage capacity costs around US$50 to 60, as compared with $21.90 per acre-feet for American reservoirs of the same type. But conservation of water by means of reservoirs is a necessary step towards the full development of Taiwan's industry and agriculture. It is, therefore, a matter of absolute necessity to keep the reservoirs in good condition once they have been built. Owing to the easily eroded geological formations of the river basins, silting of reservoirs is another vital question to be dealt with. Even with a trans-valley reservoir like the Coral Lake, where the inflow can be regulated for its silt content, it has been reported that a total of 13,400 acre-feet or 10% of the original capacity has been silted up in a period of 13 years.

Soil erosion in Taiwan not only reduces the storage capacity of reservoirs, but also reduces agricultural production and handicaps flood control work. Owing to excessive forest cutting just after the war, more coarse particles were carried down by the canal waters and finally deposited on crop fields. Compared with 1938, the 1948 unit-area production dropped considerably rice, 32.8%; sugar cane, 47,4%; and sweet potato, 30.5%. The silting up of riverbeds and irrigation canals also brought much trouble and necessitated heavy expenditure for strengthening river embankments and restoring canal sections. According to the Provincial Forestry Administration, approximately 64% of the total area in Taiwan, or nearly 5.6 million acres, consists of mountainous regions reserved for forests. Only 86,000 acres, or a little more than 1½%, is denuded or under cultivation; and 355,000 acres or 6.3%, is covered with rocks and cliffs. This condition seems to be not very serious. But according to another report (Zehugraff: Forest condition in Taiwan), deforested areas totaled 372,000 acres, and 727,000 acres of steep slopes remained unprotected in the year 1951. This is certainly a matter worth immediate attention.

Another problem in water conservation has been brought up by the recent development of ground water irrigation of the sugar plantations of the Taiwan Sugar Corporation. Ground water had been used for irrigation for a long time, but with small and scattered wells the drawdown of the water table had never been a serious question. After the installation of deep wells and capacious pumps by the Sugar Corporation, a drawdown of 30-70 ft. is not uncommon, which certainly diminishes both the head and flow of neighboring small wells. How far the underground reservoir can be emptied without serious deficiency in supply or undue cost for pumping, is a matter requiring careful study with due regard to methods of recharging. The use of return flow from irrigated areas, especially where the method of continuous flow is practiced, not only conserves a great portion of valuable water but also facilitates a good drainage system for that area. An experiment has been conducted at the Taiwan University with the result that the continuous-flow method of irrigation produces nearly the same amount of rice as the rotative method but requires an additional 35% of water as compared with the latter. In other words, the rotative method can save 26% of water without causing a reduction in production. The benefits resulting therefrom do not stop here. If the same irrigation area is maintained, the reduction of flow would also mean the diminution of seepage losses in canals and ditches. If the saving of water means the increase of irrigation area, or, in the case of three-year-rotation areas, the increase of rice irrigation, the losses of conveyance per unit area would also be reduced. The loss of fertilizer by the continuous method is self-evident; but it is not a constant loss, as one stream flow may serve a long strip of land of different areas. The continuous-flow method has the advantage of requiring less attention or labor for the division of flow by rotation and less structures for regulating and dividing the same. If the deep percolation could be recovered in the lower stretches of land, the loss would be reduced to a minimum.

Losses of water along conveyance lines are usually high in Taiwan, and in many systems where a long diversion canal is necessary the loss due to seepage will not return to the same system. One remarkable instance was a small canal leading from a filtration gallery which lost 30% of its flow over a distance of only 21 miles. The largest canal system in Taiwan, the Chia-Nan Canal, was found to lose 21% of its water in its main and 38% in the lateral and sublaterals by seepage alone. Lining with concrete improves the condition considerably, and in certain cases where gravel layers are encountered along the canal, even closed conduits may be economical.

III

How far conservation schemes can preserve the water resource and prevent the occasional damages by excessive flood flows, still remains to be disclosed by surveys and studies. To what extent the development of water resources can benefit Taiwan is a question which cannot be answered until some years later. But one thing is certain: the improvement of existing systems, especially the reduction of losses and wastes, is more beneficial than the creation of new projects. For instance, the Chia-Nan Canal controls about 340,000 acres of land with two sources of water supply, one from the Coral Lake, and the other from the Cho-shui Creek. Owing to insufficiency of water supply, a special crop rotation system has been introduced to divide the benefit equally among the landowners. The system prescribes that only one-third of the area, or 113,000 acres, may receive full flow for rice irrigation. Another one-third is to be used for sugar plantation for one and a half years, and the remaining one-third for dry cropping for half a year. Both sugar cane and rice are essential to the economy of Taiwan, but dry crops like potatoes are not. If the half-year dry cropping in every three years be converter to rice plantation, it would require nearly a 100% increase in flow during a half-year period of rice season. This has been planned by building a reservoir at the source of the Coral Lake, the Tseng-wen Creek, which might supply a flow of 350 to 1,750 cubic feet per second from May to September each year. The cost of the reservoir was roughly estimated to be US$33,000,000. On the other hand, the reduction of conveying losses by canal lining or remodeling, say 30% of the diverted supply; the increase of the irrigated area by means of rotative method say another 30%; and the reclaim of returned flow, say 40% - these might help to make up the efficiency. The second proposal, however, requires a thorough study and a detailed estimate from the standpoint of economy.

Of the total cultivable area of 2,170,000 acres, about 243,000 acres will be assigned for sugar cane plantation which requires one and a half years for a crop, or 365,000 acres per annum. The remaining 1,800,000 acres, if given full irrigation, can produce two crops of rice in the same period, or 3,600,000 acres of single crop each year. In 1951, only 1,910,000 acres of single crops was reported, and not much increase can be expected this year. It is estimated that about 660,000 acres of cultivated land can be converted to perennial irrigated fields. If this could be done, it would mean 3,200,000 acres of rice crop annually. According to last year's production of 1,700 lbs. of rough rice per acre per crop, this would mean 2,490,000 metric tons. The proposed 16 multiple-purpose reservoirs may store up enough water for 800,000 additional acres, if the duty of water can be increased to the Indian standard. The remaining 400,000 acres may be irrigated by other reservoirs as well as the water saved from improved canal systems. All in all, the water required for irrigation would not be more than 120 inches annually, or, on the average, about 25% of the total water resources. This is certainly not an extravagant proposal as compared with the estimated 19% of water resources already developed.

The rice crop in Taiwan should be greatly increased according to the average of 1931-1940, which was 2,200 lbs. or one metric ton per acre. The best record was that of the year 1938, when the per acre production was as high as 2,570 lbs. In that year, 389,334 metric tons of chemical fertilizer was used for a crop area of 1,510,000 acres, or 570 lbs. per acre. Accordingly, with fully developed irrigation Taiwan can be expected to produce an annual rice crop of 3,200,000 metric tons with the application of about 800,000 metric tons of chemical fertilizer.

The manufacturing of chemical fertilizer requires much electricity, which can be developed hydraulically in Taiwan. In the three-year period of 1937-39, an extensive survey for waterpower was made on this island and the 143 new sites discovered were estimated to be capable of producing three million kilowatts. With the addition of the 32 sites already developed and under construction, the grand total came up to three and a half million. No estimate of the cost had been made, and there were quite a few sites where the development of power might be uneconomical. This showed, however, the high potentiality of hydroelectric power per unit area on this island. Further investigation will disclose the possibility of ultimate development, which may be partly carried out in multiple-purpose projects. Granting that half of the total estimated power could be developed, this would mean approximately 1,800,000 KW. In proportion to the 1951 output, it would reach 8.5 billion kilowatt hours. If 800,000 tons of fertilizer consumes an average of 2,000 KWH per ton or 1.6 billion KWH, there still remains 7.3 billion KWH for other industrial development.

When and how these ultimate goals can be reached is a question that can be answered only after both extensive and intensive studies have been made, A few suggestions for immediate action are given below:

(a) It is necessary to start basin-wide planning as soon as possible. This should be done by the joint efforts of the government agencies, including the Water Conservancy Bureau, Taiwan Power Company, Department of Agriculture and Forestry, Forestry Administration, Highway Bureau, and all other agencies related to agricultural and industrial production. Three basins may be given first priority of study and planning, viz., (1) the Cho-shui Creek which controls about 639,000 KW of potential and developed power and about 380,000 acres of irrigated or irrigable land; (2) the Ta-chia Creek which has been planned to develop a maximum hydroelectric power of about 460,000 KW and to irrigate an area of 62,000 acres; and (3) the Ta-ko-ken Creek which might develop 153,000 KW of power and irrigate an area of 156,000 acres. These three creeks comprise a total of more than one third of the potential power and more than one fourth of the cultivable land on this island. A thorough study and wise planning of their basin development will solve over 30% of the economic problems in Taiwan.

(b) Hydrological and geological data must be collected more extensively and with greater accuracy. They must be properly analyzed and preserved. Lots of old data have been lost or destroyed during the change of hands. It is a pity that during the Japanese occupation, hydrological data were collected by various agencies without a common aim. For instance, a station recording river flow for power development cared very little about floods while one installed by river workers usually neglected the minimum flow. Rain gages, too, were neither observed at the same hour of the day nor kept constantly at the same spot. Also, very few recording gages were installed for rainfall and river stages. In a word, all data available at present are either insufficient or unreliable, especially those regarding flood peaks, as no reading could be taken during typhoons, which were the cause of the worst floods. In order to be useful and exchangeable, each kind of data collected by different agencies must be based on a standard procedure.

(c) Hydraulic and other engineering laboratories, including field plot experiments for irrigation, soil conservation, etc., must be greatly reinforced or expanded in order to meet the requirements of basin development and improvement of irrigation methods and structures. The one mentioned about the difference of crop production between continuous and rotative methods of irrigation is a good example of how experiment may answer some important questions. Young engineers must be trained to find facts to support their proposals and to improve old methods or procedures.

To conclude, I should like to point out that the most vital question in Taiwan is: How can we conserve our water resources so that their development may be economical and efficient and help to promote the well-being of the people?