1

Chapter 8

Climate, droughts and water resources

Chapter 8

Climate, droughts and water resources

1Purpose

PART A. Where is the evidence?

2Historical background

2.1Ancient and modern prophesies

2.2Hydrologists of ancient Egypt

2.3The hydrological explorers 1950 - 1970

2.4The growth of applied hydrology in South Africa

2.5The present position

3Early studies of climate change in South Africa

3.1Department of Agriculture (1925)

3.2Kokot (1948)

3.3Tyson (1987)

3.41988 to 2001

4Anomalies in climate change scenarios

5Characteristic reversals

6Conclusions to Part A

7Credibility gap

Part B. Principles of water resource analyses

8Introduction to Part B

8.1What is a drought?

9Water requirements vs available resources

10Environmental concerns

10.1Equilibrium state

10.2Water allocation for environmental conservation

11The hydrological database

12Long term persistence

13Generation of synthetic sequences

14Analytical procedures

14.1The yield-capacity-risk relationship

14.2Evaporation suppression

14.3Supply development vs demand management

14.4Appropriate analytical techniques

15Firm yield – an analytical cul-de-sac

16Defining the objective

17Risk assessment

18Optimisation procedures

19Development of operating rules

20Decision support systems

21Research

22Technology transfer

23A look into the future

24Conclusions to Part B

25Finally

26References and further reading

Purpose

The principal purpose of this chapter is to determine whether or not there has been an increase in droughts that could be attributed to unnatural global warming during the past century, and whether there is likely to be an increase in the foreseeable future.. (Part A). The second purpose is to introduce advanced analytical methods that can be applied to optimise water resource development and operation in the presence of large year-to-year and decade-to-decade variability of river flow and diminishing resources. (Part B)

Droughts have been the subject of the most extensive investigations in the history of mankind, dating back to biblical times. In South Africa droughts have been a matter of national concern for more than a century. Analytical procedures for water resource development have been in use for more than 30 years. Advanced procedures required to meet future use as our water resources approach the limits of exploitation, are presented in this chapter for the first time.

Figure 1. Broad regions of sub-equatorial Africa that have similar river flow characteristics

PART A. Where is the evidence?

1Historical background

1.1Ancient and modern prophesies

... were all the fountains of the deep broken up, and the windows of heaven were opened. And the rain was upon the earth 40 days and 40 nights. Genesis, 6, 11-12

Behold, there came seven years of great plenty throughout the land of Egypt - and there shall arise after them seven years of famine. Genesis, 41, 29-30

Changes in temperature and precipitation regimes in future, particularly in respect of extreme drought and flood conditions, will have profound effects. Tyson 2001

1.2Hydrologists of ancient Egypt

The ancient civilisations did not develop in the mild climates of Europe with their plentiful water supplies but along the major rivers in the arid climates of the Middle East. The annual flooding of the fertile floodplains of the NileRiver was the foundation on which the ancient civilisations of Egypt were built. They must have appreciated the need for developing a predictive ability based on regular observations of river levels. The earliest evidence of routine hydrological observations is the regular horizontal engravings on a stone wall on an island at Aswan in the NileRiver. These were used for measuring the water levels in the river some three thousand years ago.

There can be little doubt that the Joseph’s biblical predictions (circa 1400 BC) of seven years of plenty followed by seven years of drought were based on these observations. I demonstrated this in an earlier chapter of this report.

The Temple of Edfu dates back to 240 BC. A series of underground corridors leads to a chamber connected to the river. The temple priests kept records of annual flood levels and probably developed a model relating the rate of rise of the river level early in the season to the maximum level reached later. Flood warning systems were obviously important. The flood levels in the NileRiver rose slowly enough for upstream observers to row down the river to warn those further downstream.

Kom Umbu was built in Roman times. There is a large well in the courtyard with a spiral staircase winding down its sides. The bottom of the well is connected to the river. An annual tax was levied on farmers whose land benefited from inundation by the fertile floodwater. The tax collectors developed a simple model relating the number of steps under water in the well, to the farms that were inundated. By counting the number of steps under water each year, the tax collectors could determine which lands were inundated and therefore taxable.

In 641 AD - more than a thousand years ago - an architecturally beautiful water level gauging structure was built on RoddaIsland in the NileRiver at Cairo. The record from the Rodda Nilometer is the longest available hydrological record in the world.

1.3The hydrological explorers 1950 - 1970

In 1950 the civil engineer RE Hurst analysed 1080 years of data from the Rodda Nilometer recorded during the period 641to 1946, which he used to determine the required storage capacity of the proposed new Aswan High Dam.He found an unexplained anomaly in the data. He then analysed other long geophysical records, where he found the same anomaly. These were sediment deposits in lakes (2000 years), river flow (1080 years), tree rings (900 years), temperature (175 years), rainfall (121 years), sunspots and wheat prices. This anomaly became known as the Hurst Phenomenon, or Hurst’s Ghost.

It is of particular relevance to this report, to note that this anomalous behaviour based on over a thousand years of continuous records, is obviously related to climatic perturbations. It is present in sediment deposits in lakes, river flow, tree rings, temperature, rainfall, sunspots and wheat prices. This raises the immediate question – why was this information completely ignored in climate change theory?

Other important developments followed in rapid succession. The Harvard World Water Programme (1950-1956) was the first large-scale hydrological research effort.

The electronic computer was developed during World War II to break the secret Enigma code used by the German forces. Soon afterwards IBM developed commercial mainframe computers. They appreciated that one of the principal future applications of electronic computers would be in data storage and processing. They also identified collectors of hydrological data as potential clients. The Department of Water Affairs and Forestry was one of the earliest users of IBM equipment in South Africa. Their IBM computer occupied three air-conditioned rooms and was far less powerful than today’s laptops.

IBM employed three mathematicians to promote the use of mathematics for the analysis of hydrological data. From then onwards, through to the present time, professional mathematicians and statisticians have been closely involved in water resource studies.

Research was directed towards developing mathematical descriptions of the properties of river flow that could be used directly for determining the storage capacities of dams required to sustain given demands. The first assumption was that the annual river flows were independent of one another. This is similar to successive numbers generated by a roulette wheel. It was also assumed that the data would be normally distributed. However, they soon ran into trouble when they attempted to replicate the Hurst phenomenon. The next step was to assume that a relationship existed between the flow in one year and that in previous years, which is described as persistence. One of the problems was that the records available at that time were too short to determine this relationship with confidence.

A sense of frustration is evident in the many research papers published at that time. Here is a selection of some of the more pertinent observations. Wallis, Mandelbrot and O'Connell were IBM mathematicians.

Lloyd (1967) referred to the Hurst phenomenon and commented that either the theorists’ interpretation of their own work was inadequate or their theories were falsely based. Mandelbrot and Wallis (1968) introduced the terms ‘Noah Effect’ to describe the fact that extreme precipitation can be very extreme indeed, and ‘Joseph Effect’ to describe the fact that a long period of high or low precipitation can be extremely long.

Yevjevich (1968) stated that attempts at long-range forecasts of water supply based entirely on meteorological processes had misdirected research and raised false expectations. Wallis and Matalas (1971) noted that there was a tendency for high flows to follow high flows and for low flows to follow low flows. This was referred to as hydrologic persistence and was attributed to storage processes in the atmosphere or in the drainage basin, either surface or subsurface.

Yevjevich (1972) commented that one of the earliest deterministic methods used in hydrology was the application of the concept of almost-periodic series to various hydrological sequences in search for their hidden periodicities. However, their extrapolation as the prediction of future events represented one of the most spectacular failures of past hydrologic investigations.

Wallis and O’Connell (1973) maintained that the presence or absence of long-term persistence could radically alter the expected value of reservoir design storage and hence the estimate of the firm yield. Finally, Klemes (1974) commented that ever since Hurst published his famous plots for some geophysical time series, the classical Hurst phenomenon continued to haunt statisticians and hydrologists, and that attempts to derive theoretical explanations from the classical theory of stationary stochastic processes have failed. (Emphasis added.) Now, more than 30 years later, this observation remains valid. As shown in earlier chapters of this report, the hydrological processes are not stationery. This non-stationarity has to be accommodated in advanced water resource analyses.

In retrospect, one can only admire the insight of these early hydrologists. From here onwards hydrologists lost their way. The problem arose from the assumption that persistence was present and was cyclical in nature. This being so, it should have been relatively easy to determine the relationship between one observation and immediately previous observations. When they failed to detect appreciable dependence, they assumed that multiyear persistence was not hydrologically meaningful, and could be safely ignored. They should have taken more notice of the 1973 comment by Wallis and O'Connel above. Perhaps they misinterpreted Yevjevich’s 1972 comment, and confused the undeniable presence of persistence, with the erroneous assumption of continuous cyclicity.

1.4The growth of applied hydrology in South Africa

Professor Desmond Midgley, the doyen of applied hydrology in South Africa, followed a different route. He established the Hydrological Research Unit at the University of the Witwatersrand in the 1960’s. He and his team used deterministic methods based on conceptual models, rather than the statistical-mathematical models that overseas hydrologists were attempting to develop.

In the mid-1970s, hydrologists in the Department of Water Affairs encountered the same problem that Hurst had observed 25 years earlier. There were far too many periods when restrictions had to be imposed on the water supply from the Vaal and other major rivers.

It became clear that the reservoir capacity-yield model then in use in South Africa was deficient, and that this was probably due to assumptions regarding the river flow characteristics. I assembled a team of hydrologists to examine the assumptions relating to the properties of the river flow sequences used in storage capacity-yield analyses. Each analyst used a different mathematical model, while I used simple graphical methods. The mathematical models did not provide any insight, but graphical analyses showed that there was a very clear 20-year periodicity in the data and that this was the cause of the difficulty. I published my findings in 1978 in a Department of Water Affairs’ technical report TR80 titled Long range prediction of river flow – a preliminary assessment. The graphs showed that there was a clear pattern in the accumulated departures from the record mean values. These were quite different from random deviations.

I continued my studies along these lines. As the years passed, the periodicity in the VaalRiver data approached the 95 per cent level of statistical significance required in many engineering applications. My paper Floods, droughts and climate change was published in the South African Journal of Science in August 1995. I detailed my analytical methods and also referred to the Hurst phenomenon. I concluded: The acid test that will demonstrate whether or not the 20-year periodicity continues is at hand. If the drought is broken by widespread rainfall during the next two years it will surely be conclusive.

Four months after the publication and 22 months after its submission, severe floods occurred over a wide area of southern Africa. Lives were lost and the drought was broken. The periodicity in the VaalRiver reached the 95% confidence level confirming my predictions and my model.

During the 1980s another school of thought developed in South Africa. Water resource analyses were based on the direct statistical analysis of river flow data. Versions of the auto-regressive, moving average (ARMA) statistical model were used to describe the process. Multiyear periodicity was ignored on the basis that it did not have a meaningful effect on the reservoir storage capacity-yield relationship. This was contrary to the Department of Water Affairs’ research findings.

1.5The present position

Despite all these studies by many researchers and research institutions based on huge volumes of reliable hydrological data, a small group of climate change advocates have succeeded in convincing the South African government authorities that global warming will pose a severe threat to South Africa’s future water resources.

On closer examination it will be found that these alarmist predictions are based on uncalibrated and unverified global climate models. These models are not based on the wealth of real world hydrological data but on cold climate, proxy data derived from tree rings and ice cores. They do not accommodate the all-important anomalous behaviour first reported in the biblical quotations, and confirmed by Hurst and many hydrologists since then.

No engineering hydrologist in water resource modelling or analyses would dream of developing a computer model that was inherently incapable of being calibrated or verified, and was based on proxy data in any form.

As will be described below, there were several investigations during the past century with the specific objective of determining whether or not there have been notable changes in the South African climate. No such changes were found.

During the past three years I have undertaken a thorough analysis of a large and comprehensive hydrometeorological data set. The results are reported in earlier chapters. I was unable to detect any anomalies in the data that could be attributable to unnatural climate change.

During 2004 I mentioned my difficulty in extensive enquiries addressed to international experts in water resource analyses. Their responses were unanimous. The postulated effects of climate change on water resources were no more than untested hypotheses and could be safely ignored.

While climate change advocates claimed consensus in their findings, there is another consensus among an equally important group of applied scientists who maintain that climate change theory is little more than an untested (and unverifiable) hypothesis. This is a gap that will have to be addressed by climate change advocates.

2Early studies of climate change in South Africa

There were a number of climate-change related studies in South Africa during the past century. The following is a summary of the studies most relevant to this chapter.

2.1Department of Agriculture (1925)

In 1925 the Department of Agriculture published a detailed report The great drought problem of South Africa. The report was presented by a five-person commission of enquiry appointed by the government in 1920. The introductory chapter commenced:

An Accumulative Evil

In September, 1920, the Government appointed a Commission to inquire into the best means of avoiding losses by drought, and, based on considerable investigations which necessitated travelling over a large part of the country, the Commission presented its report in October 1922.

The present chapter goes directly to the root of the trouble, and at the outset must in its appeal arrest the attention of every South African.

…the Commission states that two points seem firmly established: firstly that a large portion of South Africa was dry long before (settlers from Europe) arrived, as evidenced by the name “Karroo” and by the highly specialised drought-resisting flora of that region; and secondly, that since then enormous tracts of the country have been entirely or partly denuded of their original vegetation…

The simple unadorned truth is sufficiently terrifying without the assistance of rhetoric. The logical outcome of all is ‘The GreatSouthAfricanDesert’ uninhabitable by man.

This is the first reference to the desertification of the sub-continent – a recurring theme that continues to the present day. This will be dealt with in a subsequent chapter of this report on climate and environmental issues.

The Commission’s report continues.

Rainfall

When seeking the cause of the ever-recurring droughts …the question of rainfall is of first consideration. And no evidence has been brought forward to prove that the average rainfall in South Africa has changed during recent historic times. Variations occur and there are good and bad years, but there is no definite tendency traceable of either an upward or downward direction.

Having given a great deal of time to investigating whatever records were available, even to studying the record of nature shown in the width between the marks or rings made every year of growth on our old indigenous trees – evidence of periods when the rate of growth was greater than at other times due to variations of rainfall and temperature – the Commission could find nothing to indicate that there has been any appreciable alteration in the rainfall of South Africa, and is satisfied that the average now is no different to what it was a generation ago.