عنوان مقاله [English]
Palaeoflood hydrology is the reconstruction of the magnitude and frequency of recent, past, or ancient floods using geological evidences (Baker et al., 2002). The term "palaeo" has contributed to the general misconception that palaeoflood techniques are only used for estimating very old floods over geological timescales. However, most palaeoflood studies involve the study of the last 5000 years with an emphasis on the last millennium, or even the last100 years in ungauged basins (Benito et al., 2005).
J. Harlem Bretz (Bretz, 1929) was the first scientist to use geological evidence intensively to elucidate information about past floods during his studies of the pathways followed by the cataclysmic outburst floods from Pleistocene Lake Missoula. However, the term and concepts of palaeoflood hydrology were formally introduced by Kochel and Baker (1982). They emphasized on slack-water deposits (SWD) and palaeostage indicators (PSI) as geological, geomorphological and geobotanical evidences for reconstruction of ancient floods. These indicators represent the high stage of the flood and provide the best natural record of large flood magnitude. Slak-water deposit sites include back-flooded tributary mouths, caves and alcoves in canyon walls, channel expansions where flow separation causes eddies, and overbank floodplain deposits. Ideal palaeoflood sites preserve multiple flood stratigraphic records, which can be separated into individual flow events using sedimentological criteria (Baker, 1987 Benito et al., 2003).
Over the last 20 years, palaeoflood hydrology has achieved recognition as an interdisciplinary branch of geomorphology and hydrology. In addition it has seen major advances, which can be partly explained by the interest generated by global climate change and its effect on river system dynamics. Several studies, in fact, deal with the chronology of ancient floods with respect to climate changes over the past millennia, thus allowing for a better understanding of flood phenomena while taking into account changing climate environments (Laurent, 2004). Unfortunately, palaeoflood hydrology has been not interested in Iran however it is an emergence science for this country. Occurrence of 3 catastrophic floods in the Madarsoo River during 2001 to 2004 and incapability of traditional hydrologic methods to estimate of peak flood discharges, caused the authors considered to necessary of palaeoflood hydrology as a way to understanding of flood phenomena in the country (Hosseinzadeh et al., 2006 Hosseinzadeh, 2008). This paper deals with the using of slack-water sediments and hydraulic calculations for the reconstruction of the high-magnitude flood records at 3 sites in a small sub-basin upstream of Daroongar River over the last centuries. The basin area with 80 km2 is located in the northeastern of Iran across the border between Iran and Turkmenistan.
At three study sites the river is confined by bedrock walls, as the river cuts west-east through the Cretaceous thick layers limestone of TIRGAN in the kopet Dagh Mountain. The river channel is a deep bedrock canyon, although at the wider gorge bends gravel bars occur. In the case of the Daroongar River the protected sites mainly include caves and alcoves in the canyon walls. Three main flood deposit sites were studied along the Shamkhal canyon. The bedrock canyon allows the assumption that little to no change in the shape of the canyon throughout the late Holocene occurred.
The main objectives of this paper are: (1) reconstruction of the catalogue of major flood events using the stratigraphic record of slackwater flood deposits, (2) study of palaeoflood hydraulics associated with these floods with estimation of flood peak discharges, (3) analysis of palaeoclimatic conditions related to the palaeoflood periods derived from stratigraphic data. primary study to finding of slackwater deposits was using the aerial photographs with 1:20000 scale and satellite Images from google earth website. Then we walked through the canyon to mark the favorite slackwater deposit sites for further studies. In the next step, 3 sites were selected at the beginning of the canyon and filed works concentrated on these sites to detailed study of sedimentology, stratigraphy and channel geometry. During the period October 1-3, 2010 the flood deposit profiles were exposed by digging and cutting. This enabled a visible exposure of the entire section of the SWD relic from top to bottom, as well as correlation between sedimentary units. The sequence of deposits exposed in each profile was separated into flood deposits associated with flood events using well-established sedimentological criteria (Baker, 1987 Benito et al., 2003) (Figs. 7, 8). This sedimentological separation enabled the reconstruction of the stratigraphy at each site. Each unit of the deposits at each site was documented in detail and sampled for optional future laboratory analyses.
The height of slack-water sediments in the selected sites can be extrapolated to the mainstream for a conservative estimate of peak paleoflood stage. Once peak flood depth is known, the slope-area method [Dalrymple and Benson, 1967] can be used to indirectly estimate paleoflood discharge. Field data necessary for slope-area calculations include flood depth, channel cross sections (Figure 5), estimates of floodplain and channel roughness, and estimates of water surface slope. Paleoflood discharges computed using the modified Manning equation:
Where is roughness A., is cross-sectional area (m2),R is the hydraulic radius, and S is the water surface slope. Roughness estimates varied between 0.035 and 0.04, near the values reported for bedrock streams of 0.04 to 0.05[Chow, 1959].Channel bed slope was substituted for S hence the values are probably low.
The height of slack-water sediments in the selected sites can be extrapolated to the mainstream for a conservative estimate of peak paleoflood stage. Once peak flood depth is known, the slope-area method can be used to indirectly estimate paleoflood discharge. Field data necessary for slope-area calculations include flood depth, channel cross sections estimates of floodplain and channel roughness, and estimates of water surface slope.
The survey used a total station with a laser rangefinder to form a series of 10 cross-sections every 40 m in average along the study segment of the upper Shamkhal canyon. The measurements include all geometric parameters of the channel and the entire canyon of the, such as width, depth, present bed river, and gradient hydraulic parameters, such as roughness coefficient, and sinuosity and the elevations and locations of all PSIs, such as driftwood lines, SWDs.
Results and Discussion
This site is an alcove at the left bank of river in the beginning of shamkhal canyon. four depositional sequences were found at this site which included 20 flood depositional units. these sequences are separated by none-flood stratums which characterized by human activities or extreme bio-turbation . These flood units consist of 50-60 percent silt, 20-30 percent fine to medium sand and about 20 percent clay. Sediments featuring mainly diffused lamination flow structures. The contacts usually consist of mud cracks, clay accumulations and charcoal or organic materials.
This site is been formed in the end of river bend at the top part of an alluvial terrace approximately 3 m above the bed river of the Daroongar. Depositional units were found at this site, 12 of which correspond to flood deposits. The flood deposits consist of fine and medium to coarse sand, featuring diffused lamination, with many charcoal pieces mixtures with some units. The highest layer of slackwater deposits had been covered by paleosoils, rock debris and slop-wash materials. The contact lines are characterized by color and grain size changes.
This site is an alcove in the steep-side gorge which is originated by faulting activity. Four depositional units were found at this site, two of which corresponds to flood deposits. The flood units consist of coarse sand to silt, featuring diffused lamination and bio-turbation.
Slope-area calculations at the site indicate that the peak discharge reconstructed for floods associated with deposits at the top of site 1 is about 350 m^3/s -1 and about 340 m^3/s -1 for the top of site 2. Calculated peak paleovelocities for these floods are between 5 and 6, respectively. Peak discharge calculations after 20% increases the flood surface on the top of slack-water deposits resulted up to 500 m^3/s in both sites. Based on another flood mark on the steep-side canyon walls, peak flood discharge is been estimated about 700 m^3/s.
Paleoflood hydrology is an important topic in fluvial geomorphology, which is an emergence science for estimation of large flood magnitude and frequencies in Iran. The results of this paper shows existence of large errors in peak discharge calculations when the traditional methods is been used. The peak discharge in the Daroongar gage station has been recorded 202 m^3/s for an area with 942 Km^2, developed for study area about 65 m^3/s based regional analysis and SCS methods. We estimated the peak discharges about 350 m^3/s based on slack-water deposit hights and more than 500 m^3/s with using other paleostage indicators. Using the paleflood methods in the large rivers of Iran can improves the predicting of peak discharge methods. Duo of short period records in the gage stations, paleoflood methods will be demonstrate confident results for flood risk assessment and environment planning projects.