1 An Introduction to Forest Hydrology
L. Bren *
The University of Melbourne, Creswick, Victoria, Australia
1.1 What is Forest Hydrology?
Forest hydrology is the study of the structure and function of watersheds and their influence on water movement and storage. In its purest form it is a quantitative discipline underpinned by conservation of mass and energy in connected, continuous media. However the application of such âpureâ theories is rendered difficult by the variations, both of inputs across space and time and in the properties of materials comprising the watersheds. Such difficulties are the stuff of forest hydrology.
In writing an overview of the discipline, one is struck by the vastness of the publications across what might be described as âforest hydrologyâ. These encompass theory, observations, methodologies, processes, results and political advocacy. The scale of work ranges from molecular to effectively the size of the earth. Interests may be in the science, economics or politics of land-use management. Forest hydrology grades into the wider disciplines of meteorology, geology, hydrology, forestry, soil science and plant physiology. There is a diverse and voluminous worldwide literature in the discipline.
Interest in forest hydrology dates back to three sources; the first of these was intellectual curiosity about the way the world works. The second was the observation of landholders that actions such as clearing forests often generated consistent, observable and (in hindsight at least) predictable results in streamflow and sediment load. The third was an age-old concern about the âsustainabilityâ (as we would now define it) of land uses and, in particular, of rainfall. Underpinning this was and is, of course, the socio-economic importance of streamflow to the survival of communities and some harsh experiences when rainfall and consequent streamflow was either extremely low or extremely high.
1.2 Development of Forest Hydrology
1.2.1 Historical antecedents
In practical terms, the history of hydrology dates back to the earliest civilizations such as ancient Rome since they certainly had the ability to measure flows and to manage water with canals, drainage tunnels and dams. Scientific historians note the growth of hydrological science for many centuries but usually denote the starting point as the work of Frenchmen Pierre Perrault (1608âÂ1680) and Edme Marriotte (1620â1684) in the period 1670â1680. This showed that the rainfall in the Seine Basin was entirely adequate to sustain the flow of the river (Biswas, 1970). Around 1700 the English astronomer Edmond Halley advanced the field further by providing the first quantitative estimates of what we would now call the hydrological cycle (Hubbart, 2011). Unfortunately there seems to be little information on who first formulated that key complementary idea to the rainfall â the watershed. McCulloch and Robinson (1993) suggest that the concept has been used for millennia. However the well-known scientist Cayley (1859) refined the concept of contours and slope lines and might well be viewed as an early scientific user. Examination of early dam-building projects in Australia, at least, suggest size was usually based on the size of the river feeding the dam, and that determination of the size and properties of the watershed usually came (much) later.
The emergence of forest hydrology as a sub-Âdiscipline of hydrology appears to owe much to the unfortunate victims of the guillotine in the French Revolution (Andreassian, 2004). This led to an unparalleled expansion of land clearing in France as âthe Kingâs Forestsâ were cleared for settlement. Landholders then encountered many of the same problems â erosion, flooding, streams drying up, landslips or other forms of mass erosion, and sedimentation â now encountered in developing countries. At the time, France was probably the most technically advanced country in the world. The ills and possible remedies caused much discussion in intellectual circles of post-Ârevolutionary France, although by modern standards the discussion was philosophical rather than scientific. Out of this came a view of the forested watershed as being something analogous to a âspongeâ (sometimes called the âLaw of Dausseâ after Dausse, 1842) and this oversimplification still underpins the view of non-technical citizens.
Among other things, Dausse (1842) argued that âRain is formed when a warm and humid wind comes in contact with strata of cold air; and since the air of forests is colder and more humid than that of the open, rain must fall there in greater abundanceâ. The view was then expressed that the forests constitute âa vast condensing apparatusâ. This message became codified into âtrees bring rainâ, which was a worldwide catchcry of a century ago. Interestingly, satellite measurement of air temperatures in the last decade have at least confirmed that the air of forests is colder than surrounding agricultural land because of heat loss associated with greater transpiration (e.g. Mildrexler et al., 2011), but the link to greater rainfall and condensation appears elusive and is a fertile field for future research using todayâs technology. This sort of approach can be viewed as a progenitor of more modern science applied to the same field. Subsequent chapters in this book will still explore some of the same ideas.
1.2.2 The era of hydro-mythology
In the latter part of the 19th century, views concerning the role of forests in hydrology began to become accepted and, indeed, were viewed as âconventional wisdomâ. These include âtrees bring rainâ, forests modify flooding, forests provide âhealthier waterâ, forests provide increased dry-Âseason flows and that forests reduce erosion. A century and a half later, such statements would be viewed as âpartly trueâ, âgeneralizations, âsweeping statementsâ or âunprovenâ but are still commonly cited by the media. In this period, data started to be collected to âproveâ such statements; the concepts of experimental design, rigorous measurement and hypothesis testing were yet to arrive in the world of forest hydrology.
By the start of the 20th century there was a body of advanced thought on the role of forests in protecting watersheds and some skilled observation, but little that we would now recognize as âscienceâ. Some authors (e.g. George Perkins Marsh, 1864; Raphael Zon, 1912) were far ahead of their time and contemporaries in examining the beneficial effects of the presence of large forests on streamflow. In retrospect, their work was a seminal contribution to the developing field of forest hydrology and watershed science. With the development of forestry science, stable forest management organizations and the advent of sophisticated and reliable instruments (e.g. water level, precipitation, air temperature and solar radiation recorders), the discipline was ripe for development.
1.2.3 The era of small watershed measurement
Around the middle of the 19th century the value of hydrological data was realized. In general this took the form of periodic readings of major river levels. Although informative, it was quickly realized that with this approach it was impossible to link rainfall and streamflow except in the crudest sense, and that large rivers were both difficult to measure flow on and too complex for simple water balance studies. This led to the first true âwatershed studyâ in the Bernese Emmental region of Switzerland in 1906. In this the hydrological responses of two watersheds of 0.6 km2 were compared. These had different distributions of land use. Inferences on the hydrology of the slopes were drawn by comparison. In general, the results showed a moderating influence of the presence of forests on peak flows and a slower summertime recession from the forested watersheds (reflecting better slope storage). Measurement at Emmental still continues and the data set is a valuable asset for climate change researchers; Hegg et al. (2006) provide an overview of this project.
By contemporary standards, the early Emmental project was far from perfect. It relied on correlation between land use and outputs rather than experimental manipulation, data were sometimes discontinuous, and the project appears to have had a somewhat tenuous political existence. From this writerâs distant viewpoint (in space and time) one has to admire the work and the people that made it happen â going out to the field on horseback or on foot, measuring in wet and cold conditions, countless hours of tedious calculations using hand calculators, logarithmic tables or slide rules, laborious hand-plotting of graphs, the constant struggle to maintain and upgrade equipment, and the ever-present demand from administrators of âwhat is more data going to show you that you donât already know?â However the project did set the scene for the big advance in forest hydrology â paired watershed experiments.
1.2.4 That great leap forward; paired watershed experiments
The European experiences were not lost on a generation of US settlers, with massive efforts directed at controlling large rivers. The value of forests in protecting watersheds was explicitly recognized by the formation of the National Forest Service in 1891. However there was no clear basis of information beyond the earlier observations of George Perkins Marsh (1864) â a deficiency clearly evident to the early forestry scientists.
In 1910 the âWagon Wheel Gapâ experiment was commenced in Colorado by the US Forest Service (Bates and Henry, 1921, 1928). This was the first formal examination of the effects of forest denudation on streamflow and sediment yield. This study ran until 1926 and was the prototype of hundreds of paired watershed experiments around the world; arguably this has been the most successful forest hydrology technique. In this, a âto-be-treatedâ stream is âcalibratedâ against a âcontrolâ or reference stream. The forest on the first watershed is then altered and the effect on streamflow is determined by comparison with the flow in the âcontrolâ stream. Their conclusions were based on mean values of study variables without the benefit of a statistical treatment of year-to-year variability.
Van Haveren (1988) revisited the data set produced by Bates and Henry (1928) to ascertain whether a more sophisticated âmodernâ approach (including covariance and regression analysis) would give the same result as that of the older work. Table 1.1 summarizes his findings.
Table 1.1. A comparison between the van Haveren (1988) examination of the Wagon Wheel Gap experimental logging and the conclusions reached by Bates and Henry (1928).
Hydrograph parameter | Original conclusion | Re-evaluation result |
Average annual water yield | Increased 24 mm | Increased 25 mm |
Annual maximum daily flow | Increased 50% | Increased an average of 50% |
Date of the annual maximum flow | Advanced 3 days | Advanced 6 days (NS) |
Starting date of snowmelt | Advanced 12 days | Advanced 5 days (NS) |
The results of this analysis showed that âmany of the original conclusions stated by Bates and Henry (1928) are statistically supportable. However a few of their conclusions could not be supported statisticallyâ. The finding underlines the discipline of the early researchers doing what is now viewed as âcomputationally intensiveâ work in the pre-computer days. Study of the Bates and Henry (1928) work also shows tentative first steps in âhydrograph analysisâ â relating specific characteristics of the flow record to the land use or land-use change. This continues to be something of a specialty area in the discipline of forest hydrology today.
1.2.5 Proliferating paired watershed experiments
The success of Wagon Wheel Gap led to a large increase in paired watershed projects around the world; these can be generally classed as âdeforestation experimentsâ in which the effects of forest harvesting were stud...