paleoflood hydrology, new approach of modern geomorphology in flood hazard assessment

1. introduction current statistical methods may be unable to accurately predict recurrence intervals of rare, large magnitude floods. the usual procedure involves extrapolation from gauged hydrological records documenting 30–40 year records of observed (normally small) floods to the estimation of the quantifies and recurrence intervals of very large, rare floods. this conventional approaches become less reliable as recurrence intervals exceed the length of the data base. if anything that happened in the past can happen at any instant with the same likelihood, then history provides no meaningful information and conventional method of addressing flood risk assessment can be improved by including information on past floods. past flood information can be obtained from palaeoflood and from historical information. documentary records can provide a catalogue of the largest flood events that curried during periods of settlement, while palaeoflood investigations using palaeostage geological indicators can document the magnitudes of the largest floods over well defined periods of time (usually from decades to millennia), and provide evidence of all other events below or above specified flow stages or thresholds. for historical flood data, human observation is required, but the modern hydrological procedures employed at gauging stations do not apply. both sources are types of non-systematic information and use the same statistical analysis approach. palaeoflood and historical flood data provide a feasible solution for assessing and mapping flood risks, and planning flood-prone zones. these improved flood-potential estimates using palaeoflood and historical flood data also have significant beneficial economic and environmental implications, related to floodplain planning, management design of hydraulic structures, management of critical water-resources and environmental conservation issues. this paper is focused on introducing of paleoflood hydrology as an approach of modern geomorphology in flood risk assessment with emphases on using the technique in iran. therefore the present paper has a review substance which is based on all results throughout the world and 6 case studies in iran. 2. paleoflood hydrology: definition and background the term of paleoflood first came into prominent use in the 1970s (costa 1974; patton and baker 1977; baker et al 1979). the recordings of palaeofloods are natural, not human, and this distinguishes them from instrumental/systematic and from historical floods measurements. paleoflood hydrology is the science of the study of paleofloods, which is emerged from divers studies in quaternary geology and fluvial geomorphology. in definition palaeoflood hydrology is the study of past or ancient flood events which occurred prior to the time of human observation or direct measurement by modern hydrologic procedures. most studies involve prehistoric floods, although the methodology may also be applicable to historic or modern floods in remote areas not subject to either modern hydrologic analysis or to human observation. these indices include various effects on the landscape, sediments, or vegetation. then sources of palaeoflood data are geological indicators such as flood deposits, silt lines and/or erosion lines found along a river’s channel, valley walls and/or terraces etc. the background of paleoflood studies during the 1970s to recent years has been summarized in table 2 with a dividing the age into 3 following stages: 1. the origin of the discipline (before 1980), 2. the evaluation period and regional development (1980-2000), 3.globalization phase. the origin based on a review from baker(2008: 2) is returned to dana(1882), tarr (1892) , flint(1933), (fuller, 1917) and specially harlen bretz (1923,1928 and 1929).then in 1982 paleoflood hydrology received its name by kochel and baker (baker, 2008: 5). baker (2006, 2008 and 2013) reviews the origin and progress of the discipline in all around the world, the below records are been summarized directly from his papers with referring to table 2 in the persian section of this paper. paleoflood studies were started with an extensive program of investigations in central texas during the early and middle 1970s (baker, 1975; patton and baker, 1977). during the early 1980s, the focus of southwest us paleoflood research moved to arizona and adjacent portions of utah (ely and baker, 1985; webb et al., 1988). about this same time, a related program in paleofloodhydrology was initiated in colorado (costa, 1983; jarrett, 1990), california (ostenaa et al., 1996) and utah (ostenaa et al., 1997, greenbaum et al., 2005). then paleoflood investigations has been pursued for many years in the north–central united states (knox, 1985, 1993, 2000), northeastern us (brown et al., 2000), and limestone caves in the appalachianareas (springer and kite, 1997; aldred,2010). paleoflood studies are also finding application in the northwestern us (o’connor et al., 2003) and in canada (gottesfeld, 1996; saint-laurent et al., 2001, 2010). paleoflood investigations based on swd and psi developed to northern australia during the early 1980s (baker et al., 1983, 1985 and pickup, 1987; pickup et al., 1988 ) and followed in india during the lately 1980s to the beginning of 2000s (baker,1988; ely et al., 1996; kale et al., 1997, 2000, 2003) more recent studies have been accomplished in southern india (thomas et al., 2007; kale 2008). studies have also been conducted in nearby thailand (kidson et al., 2005)and nepal (wohl, 1995). the best settings for swd-psi paleoflood hydrology occur in southern europe. extensive work has been conducted in spain (benito et al., 2003; thorndycraft et al., 2005). more localized studies were done in greece (lewin et al., 1991;macklin et al., 2010), italy (dewaele et al., 2010), and southeastern france (sheffer et al., 2003, 2008) paleohydrological investigations in northern europe have predominately focused upon interpretations of alluvial stratigraphy in floodplain and terrace sequences. extensive work has been done in the uk (brown et al., 2001; macklin and lewin, 2003), germany (hoffmann et al., 2008), and poland (starkel et al., 2006). werritty et al. (2006) described a study in scotland. luo (1987) reported the detailed records of extraordinary floods on major chinese rivers and results used in paleohydraulic calculations of peak discharges at appropriate cross sections (shih, 1985). subsequent chinese paleoflood research has generally followed the paleoflood analytical techniques employed internationally by other paleoflood researchers (e.g., yang et al., 2000; yu et al., 2003; zhu et al., 2005). paleoflood studies in japan are summarized in papers by jones et al. (2001), grossman (2001), and oguchi et al.(2001). in the early 1990s, a detailed program of paleoflood investigations was begun in the negev desert of southern israel (wohl et al., 1994, greenbaum et al., 2000, 2001,2008 ). in iran a catastrophic flooding in the northeastern of the country caused iranian researchers looked for a new way to calculate of large flood discharges and asses the highest flood level in the channel rivers, so they used the paleoflood techniques in many bedrock canyons in the country (hosseinzadeh and jahadi toroghi 2007, 2013, hosseinzadeh et al, 2013 and 2014). the paleoflood investigations in southern africa were accomplished by smith and zawada (1990), zawada (1994, 1997, 2000), and zawada and hattingh (1994). 3. slack water deposits (swd) and paleostage indicators (psi) slack-water deposits and palaeostage indicators are used to infer the past flood stages. river channels cutting through hard bedrock or other resistant boundary materials (e.g., cemented terraces) provide the most suitable settings for reconstructing the palaeoflood record as they are the most conducive to the accumulation and preservation of flood deposits and palaeostage indicators (psi’s). slack-water deposits consist of sand and silt (occasionally gravel) that accumulate relatively rapidly from suspension during major floods, particularly where flow boundaries result in markedly reduced local flow velocities. for palaeoflood studies, a slack-water sedimentation site should be optimum for both the accumulation and preservation of the relatively fine-grained flood sediments carried high in flood flows at maximum stage. they may develop at abrupt channel expansions, in the lee of bedrock protrusions, at meander bends and at various channel-margin alcoves (in shallow caves along bedrock walls). these fine-grained deposits can be preserved in stratigraphic sequences, providing detailed and complete records of flood events that extend back several thousands of years. tributary mouth slack-water sites are among the most easily recognized in a reconnaissance study of potential palaeoflood investigations. tributary junction angles relatively close to 90 ° are optimum for facilitating deposition by backflooding up tributary mouths. however many investigators stipulate the tributary mouths are the best sites for swd accumulations, but our studies in the bedrock canyons located at the kopet-dagh mountain range show mostly sites are been informed in the caves and bedrock alcoves (hoaaeinzadeh and jahadi toroghi 2012: 92). the thickness of slack water deposits different from a few centimeters to tens centimeters based on geomorphic situation and watershed characteristics. in the north american large rivers the tick sediment layers are common but in the rivers which located in the european and iranian mountain ranges thin layers are dominated. comparisons of slack water sediment thickness in different sites show a relationship between slack water deposit thickness and duration of flooding, the deposit thickness increases by increasing of the peak discharge duration. mean slack water deposit size is partially inherited from the lithology of the mainstream upstream from the depositonal site. the grain size of sediment samples were obtained from she-hezar river in the north of iran ranges between medium to coarse sands, the mainstream originated from alam- kuh batolith with coarse grain granite and frost weathering produce a large amounts of procedures for flood transportation. in contrast the samples which collected from kopet- dagh mainstreams include of silt, clay and very fine sands because large areas in the catchment region have covered by marl and clay formations. in addition, grain size is controlled by the fluvial regime of the main river. 4. paleoflood dating methods once the discharges of paleofloods are determined along a river reach, the absolute flood chronology can be established using by one of dating methods. these methods included dendrogeomorphology, radiocarbon dating, lichenometry, thermal luminescence (tl), optically simulated luminescence (osl) and cesium137. however dendrogeomorphology is the best accurate method to precise even the year or season flooding but this technique faces with limitations in the length of the dating. radiocarbon dating is more popular because its possible to date the paleoflood for longer period of time more than to 40 thousands years. early works in paleoflood hydrology relied almost extensively on conventional radiocarbon dating for its geochronology. since the 1980s, however, spectacular advances have been made in geochronological techniques for precisely determining the ages of ancient floods. most important have been tandem accelerator mass spectrometry (tams) and osl dating(baker, 2008: 7). the first method permits the precise dating of tiny carbon-rich material that was transported by a flood and secondary one permits the dating of individual sand or silt grains that were transported in suspension by flood vent. cesium137 is also used for dating of very young deposits (ely et al., 1992; thorndycradt et al., 2005). unfortunately there are not any laboratories inside of iran to dating the samples in own country. it should be fine if large universities make enough funds to establishment of dating laboratories. 5. hydraulic modeling hydraulic flow models can be used to predict peak flood discharges when field surveys of channel geometry permit the association of palaeostages with discharges in channels. computerized procedures for hydraulic flow modeling are used to tie the elevations of the highest slackwater deposits to surveyed river cross sections. correlation of multiple swd-psi sites along a river reach is used to identify the maximum palaeostage achieved by a given flood. the early paleoflood hydrology investigations used the slope-area equations. the largest probable floods have been calculated in 6 bedrock rivers in iran by using manning formula and paleoflood stages based on slack water deposit elevations (table 2 in the persian section). important advances have accrued since 1980s in various aspects of computational hydraulics. various two-dimensional models, particularly depth-averaged approaches, are increasingly being employed for paleoflood studies (baker, 2008: 7). 6. paleoflood hydrology in iran iran is located in the particular climatic and geomorphic location which causes catastrophic flooding in many regions every year. flash floods in small catchments as tajrish(capital tehran), masuleh (a historic village in gilan) and bujan(a village in khorasan razavi) ; large floods in larger rivers as neka, tajan (northern slopes of alborz), madarsoo(national park of glestan) and ghareh aghaj basin(in zagros) are showed the flooding capability of this country. gage station records are not long enough to calculate the interval recurrences of large floods but it is possible to find many paleoflood sites in most of rivers. based on our observations in bedrock canyons in kopet-dagh and zagros mountain areas there are many paleoflood slack water deposit sites in the channels that preserved in a very good condition. however it's difficult to find confident slack water deposit sites in the humid climate of northern iran but the reorganization of past floods is probable using dendrogeomorphology (hosseinzadeh and jahadi toroghi, 2013). the comparison of large flood discharges obtained by paleoflood data with discharges calculated by statistic methods suggested large differences between these methods. 7. conclusion and prospect paleoflood hydrology was separated from quaternary geology and fluvial geomorphology in 1980s. after nearly three decades research in paleoflood hydrology, it has found a global development and many countries using the methods for assessing the flood hazard and environmental managements. one of new approach is reorganization of the origin of sediments to understanding of the paleostorm spatial distributions, especially in european and iranian watersheds which they have a very clear slack water deposit units. in other words the slackwater deposit layers at these rivers can be separated in color, grain size, mineralogy and thickness, which help to illustrate the origin of sediments. paleoflood hydrology can detect of climatic changes, explain the drainage network response to these changes and the effects of atmospheric circulation patterns on flooding for a long period of time. planning for establish of global database for paleoflood data will help researchers to find a global view of paleofloods for further studies. flood hazard studies in iran are mostly involve statistical methods by analyzing of yearly peak discharges to calculate the recurrence of large flood events. other works like watershed management projects use the experimental equations or computer modeling for hydrological analyses. since 1980s, turning of processes geomorphology to using the historic geomorphology methods has made modern geomorphology by combination of process and historic geomorphology, inside of this paradigm, paleoflood hydrology emerged from modern geomorphology. iranian academic centers and governmental organizations who work on natural environment were stopped in the process geomorphology stage, not passed to using modern geomorphology techniques, similarly in flood studies they were using the traditional hydrologic methods for analyzing, predicting and planning. it’s the time to select a new approach for flood risk analyzing in our country, the rivers are historian can help us to reconstruct the flood stages and flood discharge for a longer calendar. young geomorphologists should be better to pick-up their equipments and do filed work for getting truly data from nature to make them able for accurate predicts.