Chemical erosion in the eastern Himalaya: Major ion composition of the Brahmaputra and δ 13C of dissolved inorganic carbon

Geochimica et Cosmochimica Acta, Vol. 69, No. 14, pp. 3573–3588, 2005 Copyright © 2005 Elsevier Ltd Printed in the USA. All rights reserved 0016-7037/05 $30.00 ⫹ .00 doi:10.1016/j.gca.2005.02.033 Chemical erosion in the eastern Himalaya: Major ion composition of the Brahmaputra and ␦13C of dissolved inorganic carbon SUNIL K. SINGH,1* M. M. SARIN,1 and CHRISTIAN FRANCE-LANORD2 1 Physical Reasearch Laboratory, Navrangpura, Ahmedabad-380009, India 2 CRPG-CNRS, Vandoeuvre Les Nancy, Cedex, France (Received September 27, 2004; accepted in revised form February 18, 2005) Abstract—Major ion composition of waters, ␦13C of its DIC (dissolved inorganic carbon), and the clay mineral composition of bank sediments in the Brahmaputra River System (draining India and Bangladesh) have been measured to understand chemical weathering and erosion and the factors controlling these processes in the eastern Himalaya. The time-series samples, collected biweekly at Guwahati, from the Brahmaputra mainstream, were also analyzed for the major ion composition. Clay mineralogy and chemical index of alteration (CIA) of sediments suggest that weathering intensity is relatively poor in comparison to that in the Ganga basin. This is attributed to higher runoff and associated physical erosion occurring in the Brahmaputra basin. The results of this study show, for the first time, spatial and temporal variations in chemical and silicate erosion rates in the Brahmaputra basin. The subbasins of the Brahmaputra watershed exhibit chemical erosion rates varying by about an order of magnitude. The Eastern Syntaxis basin dominates the erosion with a rate of ⬃300 t km⫺2 y⫺1, one of the highest among the world river basins and comparable to those reported for some of the basaltic terrains. In contrast, the flat, cold, and relatively more arid Tibetan basin undergoes much slower chemical erosion (⬃40 t km⫺2 y⫺1). The abundance of total dissolved solids (TDS, 102–203 mg/L) in the time-series samples collected over a period of one year shows variations in accordance with the annual discharge, except one of them, cause for which is attributable to flash floods. Na* (Na corrected for cyclic component) shows a strong positive correlation with Si, indicating their common source: silicate weathering. Estimates of silicate cations (Nasil⫹Ksil⫹Casil⫹Mgsil) suggest that about half of the dissolved cations in the Brahmaputra are derived from silicates, a proportion higher than that for the Ganga system. The CO2 consumption rate due to silicate weathering in the Brahmaputra watershed is ⬃6 ⫻ 105 moles km⫺2 y⫺1; whereas that in the Eastern Syntaxis subbasin is ⬃19 ⫻ 105 moles km⫺2 y⫺1, similar to the estimates for some of the basaltic terrains. This study suggests that the Eastern Syntaxis basin of the Brahmaputra is one of most intensely chemically eroding regions of the globe; and that runoff and physical erosion are the controlling factors of chemical erosion in the eastern Himalaya. Copyright © 2005 Elsevier Ltd Brahmaputra, to the Bay of Bengal, exceeded that of the Ganga (Hay, 1998). The limited studies on the Brahmaputra (Sarin et al., 1989; Galy and France-Lanord, 2001) indicate that the eastern Himalaya, encompassing the Brahmaputra drainage, is eroding more rapidly than the central and western Himalaya which make up the Ganga drainage. This difference has been attributed to higher rainfall and runoff in the eastern sector (Sarin et al., 1989; Galy and France-Lanord, 2001). More recently, chemical, isotopic, and mineralogic data on sediments from the Brahmaputra watershed show that the basin is characterized by differential erosion and that the sediment budget of the Brahmaputra is dominated by material derived from the Eastern Syntaxis region (Singh and France-Lanord, 2002; Singh et al., 2003; Garzanti et al., 2003). These results are also supported by model calculations (Finlaysan et al., 2002). To gain better understanding of the relative erosion rates within the Brahmaputra basin, a detailed investigation on the chemical and isotope composition of the water and sediments was undertaken. The source apportionment of the solutes in the Brahmaputra River and its tributaries from silicates and carbonates is an integral part of this study for determining the CO2 consumption rates. Recently, several studies have been carried out to quantify the sources of solutes in the Ganga drainage (Singh et al., 1998; Galy et al., 1999; Krishnaswami et al., 1999; Dalai et al., 2002; Bickle et al., 2003). In this context, studies in the 1. INTRODUCTION The Ganga-Brahmaputra (G-B) is one of the major global river systems, ranking first in sediment supply and fourth in water discharge (Milliman and Meade, 1983; Milliman and Syvitski, 1992; Ludwig and Probst, 1998). The sediment flux from the G-B may account for as much as ⬃15% of the global sediment discharge if its bed-load equals that of its suspended load (Galy and France-Lanord, 2001). This makes the G-B system a significant component of the sediment budget of the oceans. In terms of water discharge, G-B accounts for ⬃3% of global riverine water supply (Hay, 1998 and references therein). Among these two major rivers, the Ganga has been studied in far greater detail for its chemical and isotopic composition and their impact on ocean chemistry. These studies have brought out the role of these rivers, particularly the Ganga, in contributing to Sr and Os isotope composition of seawater during the Cenozoic and on the CO2 drawdown on million-year time scales (Galy and France-Lanord, 1999; Krishnaswami et al., 1992; Palmer and Edmond, 1989; Harris et al., 1998; Blum et al., 1998; Galy et al., 1999; Richter et al., 1992; Bickle et al., 2003; Dalai et al., 2002). It is well documented that water and sediment delivery via * Author to whom correspondence should be addressed (sunil@prl. ernet.in). 3573 3574 S. K. Singh, M. M. Sarin, and C. France-Lanord Fig. 1. Map of the Brahmaputra watershed with sampling locations. Brahmaputra was sampled from Pasighat in India to Yamuna Bridge in Bangladesh. Area of study is shown in the inset. Sample numbers are indicated based on sampling in one season, and all the numbers are preceded by BR. Various subbasins are shown by dashed lines. Brahmaputra system are rather limited (Krishnaswami et al., 1999; Galy and France-Lanord, 1999). In this study, attempts have been made to determine the contribution of solutes to the Brahmaputra River System from silicate weathering and to estimate the associated CO2 consumption rates. In addition, these data represent a first attempt to document the temporal variation in the major ion chemistry of the Brahmaputra mainstream based on a biweekly sampling carried out over a period of one year, including the monsoon. Thus, these data provide better estimates of annual fluxes and seasonal variations in silicate weathering rates. 2. BRAHMAPUTRA RIVER SYSTEM: HYDROLOGY AND GEOLOGY 2.1. The Brahmaputra River The Brahmaputra River is known by different names along its course. It originates in the Kailash Mountain in the Transhimalaya and flows with a very gentle slope eastward ⬃1200 km in Tibet as Yarlung Tsangpo or Tsangpo. The Tsangpo takes a U-turn after Pai (Fig. 1) around Namche Barwa, the Eastern Syntaxis, where it makes the deepest gorge of the world and turns south to enter Arunachal Pradesh, where it acquires the name Siang or Dihang. This part of the river with the deepest gorge has a very steep slope (⬃30 m/km) causing very turbulent and rapid flow and intense physical erosion (Singh and France-Lanord, 2002). Immediately after Pasighat, the Siang River turns in SW direction and enters the Assam Plain, where it is called the Brahmaputra and flows in WSW direction as a wide and deep braided river. The Brahmaputra acquires a width of ⬃20 km and depth of ⬃35 m at some locations in the Assam Plain. The Brahmaputra turns south near Dhubri at the Indo-Bangladesh border and flows as the river Jamuna until it meets the Ganga at Arichaghat. The Brahmaputra River receives many tributaries along its course. In Tibet, the Tsangpo receives the Lhasa He (Zangbo), Doilung, and Nyang Qu (Fig. 1; Guan and Chen, 1981; Hu et al., 1982) in addition to tributaries from the northern slope of the Himalaya. After Pai, the river Parlung Zangbo (Fig. 1; Guan Chemical erosion in the eastern Himalaya 3575 Table 1. Sample details: Locations, Water discharge, Drainage, Temperature and pH. Longitude (E) Sample code River (location) Brahmaputra Mainstream Tsangpo at Pai BR-59 Siang or Dihang (Pasighat) BR-18 Brahmaputra (Dibrugarh) BR-28 Brahmaputra (Tezpur bg.) BR-65 Brahmaputra (Tezpur bg.) BR-5 Brahmaputra (Guwahati) BR-51 Brahmaputra (Guwahati) BR-73 Brahmaputra (Dubri) Brahmaputra (Chilmari)b BR 200 Brahmaputra(Jamuna bg.) Eastern Tributaries BR-14 Dibang BR-16 Lohit Himalayan Tributaries BR-20 Subansiri BR-61 Subansiri BR-24 Ranga Nadi BR-57 Ranga Nadi BR-26 Jia Bhareli BR-63 Jia Bhareli BR-75 Tipkai BR-32 Manas Biki BR-71 Manas Biki BR-34 Puthimari BR-69 Puthimari Southern Tributaries BR-10 Dhansiri BR-12 Buri Dihing BR-30 Kopili BR-67 Kopili a b Latitude (N) Dischargea, 109 m3 y⫺1 Areaa, 103 km2 deg min deg min Month & year of sampling — 95 94 92 92 91 91 89 — 89 — 20.15 51.11 51.22 51.18 44.22 44.22 59.76 — 47.83 — 28 27 26 26 26 26 26 — 24 — 4.67 30.03 36.73 36.66 11.62 11.62 1.12 — 23.35 — July 2000 Oct 1999 Oct 1999 July 2000 Oct 1999 July 2000 July 2000 Aug 1996 July 2002 59 178 323 — — 455 455 510 666 — 95 95 35.15 36.34 27 27 47.84 47.90 Oct 1999 Oct 1999 94 94 94 94 92 92 90 90 90 91 91 15.10 15.42 3.53 3.64 52.57 52.77 8.28 54.95 55.17 39.22 39.20 27 27 27 27 26 26 26 26 26 26 26 26.90 26.74 12.24 12.32 48.60 48.65 13.11 29.70 29.73 22.04 22.01 93 94 92 92 43.88 53.02 21.27 21.22 26 27 26 26 37.87 18.74 9.91 9.93 Temp, °C pH 220 246 298 376 376 384 384 455 636 — — 23.2 21.1 24.7 32.1 24.7 31.2 29.6 29.0 — — 7.9 8.4 8.3 8.1 8.6 7.8 7.8 7.6 — 22 49 13 24 21.3 18.6 8.6 8.4 Oct 1999 July 2000 Oct 1999 July 2000 Oct 1999 July 2000 July 2000 Oct 1999 July 2000 Oct 1999 July 2000 54 54 58 58 26 26 — 32 32 — 4 26 26 2 2 11 11 — 28 28 — 2 21.0 23.9 20.8 27.4 22.0 24.5 29.4 21.8 21.9 27.3 28.1 8.4 7.3 8.4 7.2 8.2 7.6 7.6 8.4 8.3 8.2 7.8 Oct 1999 Oct 1999 Oct 1999 July 2000 20 14 32 32 8 3 15 15 26.3 25.5 26.9 32.2 7.7 8.0 8.0 7.4 Rao, 1979; Goswami, 1985; and GRDC website (www.grdc.sr.unh.edu). The values are average of several years of data. Galy and France-Lanord, 1999. and Chen, 1981) merges with it. The slope of this tributary is very high and comparable to that of the Siang in this section. In the Assam plain the Brahmaputra receives the Dibang and the Lohit from the east and the Subansiri, the Ranganadi, the Jia Bhareli, the Puthimari, the Manas, and the Tipkai from the north and the Burhi Dihing, the Dhansiri, and the Kopili from south (Fig. 1). The Tista is another northern tributary of the Brahmaputra which merges with it in Bangladesh (Fig. 1). The water discharge of the Brahmaputra mainstream and many of its tributaries are given in Table 1. The major source of water in the Brahmaputra is rainfall, though meltwater and groundwater contributions are also important. In the Tsangpo in Tibet, for example, meltwater, groundwater and rainfall contributions are roughly the same (Guan and Chen, 1981). The runoff in the Tsangpo drainage is ⬃300 mm yr⫺1 which increases by more than order of magnitude, to ⬃5000 mm yr⫺1, for the Siang in Arunachal Pradesh. Runoff in the Himalayan drainage for the northern tributaries in the Assam Plain is ⬃1000 –2000 mm yr⫺1 and for the eastern tributaries it is ⬃3000 mm yr⫺1. The southern drainage is exposed to heavy rainfall and the runoff in this region is ⬃4000 mm yr⫺1. The major contributor to the Brahmaputra discharge is rainfall during SW monsoon (July to September). Monthly discharge of the Brahmaputra at Bahadurabad, based on the average of many years, is shown in Figure 2. The monthly water discharge pattern of the Brahmaputra at Bahadurabad reflects the monsoon with significant temporal variation. It varies from ⬃3300 m3/s in February to ⬃59000 m3/s in July. The discharge in February is the lowest owing to paucity of rain and less meltwater contribution. This trend is almost similar to that at Fig. 2. Monthly discharge of the Brahmaputra at Bahadurabad in Bangladesh. Most of the discharge in the Brahmaputra occurs during monsoon season, i.e., June to September. 3576 S. K. Singh, M. M. Sarin, and C. France-Lanord Guwahati in India, where biweekly samples have been collected. The Brahmaputra System drains a total area of ⬃630,000 km2. Of the total drainage, about one third is in Tibet with an average elevation of ⬃5000 m. The Tibetan drainage contributes ⬃10% of the water discharge of the Brahmaputra at its mouth. The Brahmaputra drains total area of ⬃200,000 km2 in the plains of Assam and the Bangladesh and its Himalayan tributaries occupy an area of ⬃120,000 km2 in the Himalaya. The two eastern tributaries, the Lohit and the Dibang flowing through the Mishmi Hills have drainage area ⬃50,000 km2. 2.2. Geology The drainage basin of the Brahmaputra System can be divided into five geologically and climatically different subbasins (Fig. 1). These are (1) the Tibet, the drainage of the Tsangpo, (2) the Eastern Syntaxis, the drainage of Siang and Parlung Tsangpo rivers between Pai and Pasighat, (3) the Mishmi Hills, the drainage of the two eastern tributaries, the Lohit and the Dibang, (4) the Himalaya, the drainage of the northern tributaries of the Brahmaputra in the Assam and the Bangladesh Plains (the Renganadi, the Jia Bhareli, the Puthimari, the Manas, the Tipkai), and (5) the Indo-Burmese Ranges, the drainage of southern tributaries (the Burhi, the Dihing, the Dhansiri, and Kopili) of the Brahmaputra in the Assam and the Bangladesh plains. 1. Tibet: In upper reaches the Tsangpo drains turbidites and ophiolites of the Indus-Tsangpo Suture Zone. The tributaries from the northern slope of the Himalaya drain the Tethyan Sedimentary Sequences and the gneiss zone. The tributaries from Tibetan Plateau, the Doilung, Zangbo, and Nyang Qu predominantly drain Transhimalayan gabbroic to granodioritic batholiths. The basins of these tributaries also contain evaporite deposits (Hu et al., 1982; Pande et al., 1993; Pascoe, 1963). 2. The Eastern Syntaxis: The rocks near the Eastern Syntaxis are highly metamorphosed. At its core, gneisses of the Indian Plate have been exhumed from below the Transhimalayan Plutonic Belt (TPB; Burg et al., 1998). In this zone the calc-alkaline plutons of the TPB are surrounded by quartzites, phyllites, and marbles (Burg et al., 1998). Discrete lenses of metabasites and serpentinites occur in these areas, which indicates the continuation of the IndusTsangpo Suture in the eastern section (Burg et al., 1998). These are drained by the Dibang, Parlung Tsangpo, and Lohit. 3. The Mishmi Hills: The two eastern tributaries, the Lohit and the Dibang, flow through the Mishmi Hills composed of calc-alkaline diorite-tonalite-granodiorite complexes and tholeiitic metavolcanic rocks (Kumar, 1997). It represents the eastern continuation of the TPB. The Tiding Suture present in this area marks the boundary between the TPB and the Himalaya in this section. 4. The Himalaya: The geology of the eastern Himalaya, through which the northern tributaries of the Brahmaputra System in Assam Plain, such as the Subansiri, Renganadi, Jia Bhareli, Puthimari, and Manas, flow, is similar to those of its central and western sections, which form the Ganga basin. It comprises of the Higher and the Lesser Himalaya and the Siwaliks (Thakur, 1986; Gansser, 1964). In general, the proportion of the Lesser Himalaya increases from east to west in this watershed (Singh and France-Lanord, 2002; Robinson et al., 2001). The Higher Himlayan rocks consist mainly of schists and marbles with amphiboles at some locations. In Bhutan and Sikkim, the Manas and the Tista drain through metamorphic rocks of the Higher Himalaya. The Lesser Himalaya in the Brahmaputra System drainage is composed mainly of quartzites and schists. Precambrian limestones, dolostones, shales, and quartzites along with orthogneiss bodies and dolerite sills are exposed in the Lesser Himalaya. The Siwalik is discontinuous in the eastern section of the Himalaya. It includes a thick section of Neogene molasses. Continuing uplift and deformation are evident in this section by the presence of tilted gravel terraces and steep fault scarps (Nakata, 1989). Apart from these rocks of the Himalaya the basalts of the Abor volcanics are present in the Himalayan drainage of the Siang (Jain and Thakur, 1978). The northern tributaries of the Brahmaputra in the Assam plain drain through the southern slope of the Himalaya. Only a few of them, the Subansiri, have their drainage in the Tethys Himalaya (Kumar, 1997). 5. Indo-Burmese Ranges: These ranges are made of pelagic sediments overlain by thick turbidites associated with ophiolites. The Dhansiri and the Kopili also drain the Indian basement of the Shillong Plateau and the Mikir Hills (Kumar, 1997). 3. SAMPLING AND METHODOLOGY Water and sediment samples from the Brahmaputra mainstream and its major tributaries draining between Pasighat in Arunachal Pradesh to Dhubri at the Indo-Bangladesh border (Fig. 1) were collected during two seasons: the SW monsoon and the postmonsoon. The monsoon samples were collected during the month of July, representing peak discharge, and the postmonsoon samples were collected in October (median flow). The Brahmaputra mainstream was sampled at five stations: Pasighat, Dibrugarh, Tezpur, Guwahati, and Dhubri (Fig. 1). One monsoon sample of the Brahmaputra (BR 200) was collected at Jamuna Bridge in Bangladesh. Samples were collected from the midchannels, accessing either from a boat or from road bridges. The collection procedures for sediment have been described in detail in Singh and France-Lanord (2002), Singh et al. (2003), and Garzanti et al. (2004). Soon after their collection, two separate aliquots of 500 mL water were filtered using 0.2-␮m nylon membrane Millipore filters. One of the filtered aliquots was acidified with double-distilled HNO3 for cations, trace metal, and Sr analysis, and the other aliquot was preserved unacidified for anion measurements. In addition, one sample of ⬃250 mL water was collected and stored unfiltered for alkalinity measurements. The Brahmaputra mainstream was also sampled at Guwahati (Fig. 1) at an interval of 15 days over a period of ⬃10 months to assess the temporal variations in its major ion composition. Temperature and pH of the water samples were measured at site. The water and sediment samples were brought to the laboratory for further analysis. Alkalinity was measured by acid titration; Cl, NO3, and SO4 by ion chromatography; K and Na by flame AAS; and Ca, Mg, and Si using ICP-AES. The precision of these measurements, based on earlier studies in this laboratory (Dalai et al., 2002), is about ⫾5%. Accuracy of measurements for various elements was checked by measuring dilute solutions of USGS rock standards, W-1 and G-2, and also by analyzing river waters of known elemental abundances (Sarin et al., 1992). Clay mineral analyses of selected sediment samples were done at CRPB-CNRS, France (Bartoli, 1991). Quantification of clay minerals was done by analyzing ⬍2-␮m size fraction by granulometry. ␦13C of dissolved inorganic carbon was measured using a modified VG 602D isotope ratio mass Chemical erosion in the eastern Himalaya 3577 Table 2. Major ion composition and ␦13C of DIC of waters of the Brahmaputra River System. Na⫹ Sample code Na* K⫹ Mg2⫹ Cl⫺ NO⫺ 3 SO2⫺ 4 HCO⫺ 3 TDS, mgL1 ␦13C (DIC), ‰ SiO2 ␮mol L⫺1 River (location) Brahmaputra mainstream Tsangpo (South Lhasa)a Tsangpo (South Lhasa)a Siang (Pai)b BR-59 Siang or Dihang (Pasighat) BR-18 Brahmaputra (Dibrugarh) BR-28 Brahmaputra (Tezpur bg.) BR-65 Brahmaputra (Tezpur bg.) Brahmaputra (Guwahati)a BR-5 Brahmaputra (Guwahati) BR-51 Brahmaputra (Guwahati) BR-73 Brahmaputra (Dhubri) Brahmaputra (Chilmari)c BR 200 Brahmaputra (Jamuna bg.) Tibetan Tributaries to Tsangpo Zangbo (Lhasa)a Doilung (Lhasa)a Eastern Tributaries BR-14 Dibang BR-16 Lohit Himalayan Tributaries BR-20 Subansiri BR-61 Subansiri BR-24 Ranga Nadi BR-57 Ranga Nadi BR-26 Jia Bhareli BR-63 Jia Bhareli BR-75 Tipkai BR-32 Manas Biki BR-71 Manas Biki BR-34 Puthimari BR-69 Puthimari Southern Tributaries BR-10 Dhansiri BR-12 Buri Dihing BR-30 Kopili BR-67 Kopili Ca2⫹ 396 446 387 78 106 110 78 159 86 90 107 104 77 240 248 244 61 72 86 64 52 58 69 88 79 53 32 37 28 37 48 50 49 79 50 67 50 52 62 209 191 103 100 144 140 101 119 115 111 153 168 120 752 717 500 424 540 458 378 425 395 475 396 393 433 156 198 143 17 34 24 14 107 28 21 19 25 24 — — — 14 — — 9 — — 17 11 — — 255 223 188 119 158 110 86 100 87 114 73 55 78 1670 1740 984 854 1197 1154 845 884 1005 1051 1018 1114 1060 127 125 125 126 152 189 146 123 137 140 200 155 127 185 187 150 95 128 119 91 101 102 112 106 105 105 — — — — ⫺11.5 ⫺14.6 — — ⫺13.6 — — — ⫺10.1 380 300 180 38 36 48 39 35 270 224 200 262 — — 75 60 751 558 134 88 90 73 — — 47 59 33 41 42 50 49 78 353 440 14 18 — — 82 91 780 996 141 139 82 101 ⫺12.4 ⫺12.6 79 71 137 105 101 90 124 97 59 133 99 59 62 120 94 86 80 106 78 49 109 80 24 26 33 28 32 32 27 31 27 43 36 136 108 45 37 70 60 195 148 101 291 219 323 303 158 126 235 191 284 470 416 698 551 20 9 17 11 15 10 18 19 10 24 19 — 15 — 20 — 12 7 — 8 — 19 104 95 38 34 59 33 28 148 102 143 112 849 667 550 373 682 537 994 1096 844 1877 1336 169 148 294 218 206 185 258 158 105 218 166 92 77 67 50 75 60 100 117 90 182 135 ⫺11.1 — ⫺15.5 — ⫺14.3 — — ⫺12.0 — ⫺13.6 — 288 189 137 118 205 147 104 88 63 31 40 35 226 293 99 78 230 228 158 148 83 42 33 30 — — — 3 129 85 65 64 881 1084 591 425 225 310 217 208 106 116 70 58 ⫺13.6 ⫺17.2 ⫺16.0 — a Hu et al., 1982. Chen and Guan, 1981. c Galy and France-Lanord, 1999. Na* ⫽ (Nariv ⫺ Clriv). b spectrometer following the procedure of Galy and France-Lanord (1999). 4. RESULTS AND DISCUSSION Sampling details, temperature, pH, discharge, and drainage are given in Table 1. Major ion composition of the Brahmaputra mainstream and its various tributaries are presented in Table 2, temporal data for the Brahmaputra (at Guwahati) in Table 3, and the clay mineral composition of sediments in Table 4. All samples were slightly alkaline in nature (pH 7.2– 8.6) and temperature varied from 19°C to 32°C depending on the season and on the time of sampling. In some of the following discussions, the dissolved Na has been corrected for contribution from rainwater and halites by subtracting Na equivalent to dissolved Cl from them (Na* ⫽ Nariv ⫺ Clriv). Other cations and anions are not corrected for rainwater contribution because they are insignificant (e.g., Galy and France-Lanord, 1999). 4.1. Spatial and Temporal Variability in Total Dissolved Solids The total dissolved solids (TDS ⫽ Na⫹K⫹Ca⫹Mg⫹Cl ⫹SO4⫹HCO3⫹SiO2 in mg L⫺1) in the Brahmaputra mainstream (Table 2, Fig. 3) measured in this study shows a narrow range, from 91 mg L⫺1 at Tezpur during monsoon to 128 mg L⫺1 at Dibrugarh during postmonsoon. In general, downstream variation in TDS of the Brahmaputra mainstream during the current sampling is small, and center around 110 ⫾ 15 mg L⫺1 (Fig. 3). Three data points for the Tsangpo in Tibet show higher TDS, 150 –185 mg L⫺1 (Hu et al., 1982; Chen and Guan, 1981). This is an indication that in the Tsangpo the salinity is higher and it gets diluted downstream by rainfall. The tributaries of the Brahmaputra, in contrast, show factors of 3– 4 variability in their TDS, from 50 to 182 mg L⫺1 (Table 2). The monsoon samples, as expected, are dilute in terms of their TDS compared to those of postmonsoon. 3578 S. K. Singh, M. M. Sarin, and C. France-Lanord Table 3. Temporal variation in chemical composition of the Brahmaputra at Guwahati. ⫹ Date, month/day/ year Na 10/24/1999 11/15/1999 11/30/1999 12/15/1999 12/31/1999 1/16/2000 1/31/2000 2/15/2000 3/1/2000 3/15/2000 4/1/2000 4/19/2000 4/30/2000 5/15/2000 5/31/2000 6/15/2000 7/1/2000 7/26/2000 86 141 163 171 188 229 212 220 269 195 192 99 88 107 80 95 78 90 Na* K⫹ Mg2⫹ Ca2⫹ Cl⫺ NO⫺ 3 SO2⫺ 4 HCO⫺ 3 SiO2 TDS, mg L⫺1 ␮mol L⫺1 58 108 119 130 142 150 146 149 171 131 117 71 60 75 57 73 60 70 50 54 56 56 57 66 62 64 67 61 63 52 55 58 54 154 85 67 115 193 209 228 236 246 236 233 245 208 171 120 122 119 115 192 117 111 395 588 613 658 680 711 679 686 715 632 553 457 451 455 439 1004 550 475 28 33 44 41 46 79 66 71 98 64 75 28 29 32 22 21 18 21 — — — — — — — — 4 3 1 — — — — — — 17 90 146 154 169 175 177 179 179 185 185 156 108 85 101 96 456 112 114 1005 1395 1470 1560 1634 1694 1618 1589 1676 1460 1272 1066 1081 1067 1025 1585 1238 1051 137 204 211 217 221 238 233 229 232 200 187 155 138 159 121 149 150 140 102 146 154 164 171 180 172 171 181 158 140 111 109 111 104 203 126 112 Na* ⫽ Nariv ⫺ Clriv. Table 5 is a summary of available data on TDS of the Brahmaputra along its course. These data provide information on spatial and temporal variation in TDS over about two decades. The TDS of the Brahmaputra measured in this study at Jamuna Bridge and Dhubri during monsoon (105 and106 mg L⫺1, respectively) is identical to the value of 105 mg L⫺1 reported by Galy and France-Lanord (1999) at Chilmari a few kilometers downstream of Dhubri collected during monsoon of 1996. Similarly, the TDS of 101 mg L⫺1 at Guwahati (Hu et al., 1982) during monsoon 1979 compares well with value of 112 mg L⫺1 obtained in this study. Sarin et al. (1989) measured the chemical composition of the Brahmaputra mainstream in samples collected from four sites during 1982–1983. The TDS measured at Guwahati during April 1982 was 91 mg L⫺1 compared to values of 140 and 111 mg L⫺1 measured in this study for the same month. Similarly, the value for December in 1982 was 144 mg L⫺1, marginally lower than the values of 164 and 171 mg L⫺1 in 1999. These results show that over two decades the TDS in the Brahmaputra does not show any systematic and major variations, the variability being ⬃35% or less; such interannual variations can result from associated changes in runoff. TDS of the Brahmaputra river system is generally lower than those in samples from the lower reaches of the Ganga, the Yamuna, the Karnali, and the Narayani rivers but similar to those in the headwaters of these rivers (Sarin et al., 1989; Sarin et al., 1992; Dalai et al., 2002; Galy and France-Lanord, 1999). This is most likely due to differences in drainage lithology and runoff among these basins. For example, a lower proportion of the Lesser Himalayan sedimentaries containing more easily weatherable carbonates and evaporites in the drainage basin of the Brahmaputra System can cause lower TDS in these rivers. Similarly, higher runoff in the Brahmaputra watershed can affect weathering intensity and also act as a diluent. The TDS of the Tsangpo in the Tibetan Plateau (Hu et al., 1982) is 185 mg L⫺1, the highest for the mainstream. The higher TDS in the Tibetan region is most likely due to the presence of easily weatherable evaporites in the basin (Pascoe, 1963). The chemistry of its tributaries, e.g., the Zangbo and the Doilung (Hu et al., 1982), from this region supports this inference (Table 2 and discussion). Also the lower runoff of this region helps maintain higher TDS. Comparison of TDS of the Tsangpo at Pai (Chen and Guan, 1981), before it enters the Eastern Syntaxis, and at Pasighat after its exit shows that TDS has decreased from 150 mg L⫺1 to 95 mg L⫺1 (Fig. 3). This decrease can result from changes in both geology and climate. Between Pai and Pasighat the Brahmaputra predominantly drains silicates and its runoff increases by an order of magnitude which can cause dilution and also affect weathering intensity. Downstream of Pasighat, the two tributaries, Dibang Table 4. Abundance and composition of clay in sediments of the Brahmaputra River System. Sample code River (location) Clay, wt% Clay composition BR 19 BR 29 BR 9 BR 21 BR 25 BR 33 BR 21 Brahmaputra (Dibrugarh) Brahmaputra (Tezpur) Brahmaputra (Guwahati) Subansiri Renganadi Manas Kopili 1.7 2.6 3.2 1.4 3.9 3 21 Vermicullite, illite, chlorite Vermiculite, illite, chlorite Vermiculite, illite, chlorite Vermiculite, illite, montmorillonite Illite, chlorite, Montmorillonite Illite, chlorite, montmorillonite Montmorillonite, illite, chlorite Chemical erosion in the eastern Himalaya 3579 Fig. 4. Temporal variation of various major ions of the Brahmaputra at Guwahati. All the concentrations are in ␮M except for TDS in mg ⫺1 L . One sample collected during SW monsoon has elevated concentration owing to flash flooding in the Tsangpo. Fig. 3. Downstream variation in TDS of the Brahmaputra. TDS is highest in Tibet and decreases after the river crosses the Eastern Syntaxis. In the Assam plain TDS is almost constant despite receiving many tributaries with variable TDS. More than one value of TDS at the same location represents samples from different seasons (Table 2). and Lohit meet the Brahmaputra from the east (Fig. 1). At Dibrugarh, the TDS of the Brahmaputra is ⬃35% more (128 mg L⫺1) than at Pasighat (95 mg L⫺1; Fig. 3, Table 2). This increase in TDS can be due to temporal variation in TDS, because the Pasighat was sampled in monsoon and Dibrugarh during postmonsoon, or dissolution of detrital carbonates in the Table 5. Decadal variation in TDS of the Brahmaputra along its course. Location Lhasa Pai Dibrugarh Guwahati Goalpara Dhubri Chilmari Jamuna Bridge Aricha Ghat 1. 2. 3. 4. Date TDS (mg L⫺1) Ref. Jul 1979 Jun 1980 185 187 150 107 128 101 112 91 111 144 164 102 92 147 106 105 145 105 154 1 1 2 3 This study 1 This study 3 This study 3 This study This study 3 3 This study 4 4 This study 4 Apr 1982 Oct 1999 July 1979 July 2000 Apr 1982 Apr 2000 Dec 1982 Dec 2000 Oct 1999 Apr 1982 Dec 1982 July 2000 Aug 1996 Mar 1997 July 2000 Feb 1997 Hu et al., 1982. Chen, and Guan, 1981. Sarin et al., 1989. Galy and France-Lanord, 1999. Brahmaputra during its transit from Pasighat to Dibrugarh. The decrease in abundance of detrital carbonates in sediments from ⬃ 3– 4 wt% in the Siang, the Dibang, and the Lohit to ⬃0 wt% at Dibrugarh (Singh and France-Lanord, 2002) supports the later hypothesis. This is further strengthened from the dissolved and particulate Ca fluxes of the Brahmaputra at Pasighat and Dibrugarh (see Appendix). These observations showed that in a stretch of ⬃50 – 60 km, the Brahmaputra is dissolving all the detrital carbonate present in the bed load. This interpretation can be misleading in view of different season of sampling as the Siang was sampled during monsoon and the others postmonsoon; however, the carbonate content of sediments at Tejpur, further downstream of Dibrugarh, shows no variability between monsoon and nonmonsoon sampling and therefore it can be safely assumed that detrital carbonate would not have changed between two seasons at Dibrugarh, too. The TDS of the Brahmaputra remains nearly constant from Dibrugarh to Jamuna Bridge though it mixes with many tributaries from the Himalaya (Fig. 1). The TDS of the Himalayan tributaries increase from east to west. The Subansiri, the Ranganadi and the Jia Bhareli have lower TDS as they primarily drain silicates of the Higher Himalaya and the Puthimari, the Manas and the Tipkai have higher TDS, resulting from the weathering of Lesser Himalaya having higher proportion of carbonates (Thakur, 1986; Gansser, 1964). The decrease in runoff from east to west over the Himalayan drainage can also be an additional contributing factor to this trend. TDS of the southern tributaries are comparable to those of the Himalayan drainage. Temporal variation in the TDS of the Brahmaputra River was measured at Guwahati at 15-day intervals over a period of 10 months (Table 3). TDS over the ten months showed about a factor of two variation, from 104 mg L⫺1 to 203 mg L⫺1 (Fig. 4); the range is more pronounced than spatial variations. Interestingly the minimum and maximum TDS occurred in a span of 15 days during end of May to mid-June (Table 3). The maximum in TDS seems to be a result of flash flooding in the Brahmaputra (see discussion later). The minimum during April–May is attributable to increase in glacier melt water component in the discharge. The second broad maximum of 3580 S. K. Singh, M. M. Sarin, and C. France-Lanord Fig. 5. Ternary plots of cations and anions of the Brahmaputra. Cations are dominated by Ca whereas HCO3 dominates the anion budget. Data are in ␮Eq L⫺1. 170 –180 mg L⫺1 during January–March (Fig. 4) can be explained in terms of lean flow and increased contribution from ground water. It is important to note that water discharge during monsoon is a factor of ⬃10 higher compared to that of during lean flow where as the TDS during monsoon is slightly lower (⬃0.6⫻) with respect to lean flow. This shows that the kinetics of chemical weathering and supply of solutes of river is not significantly affected despite an order of magnitude increase in runoff. most of the samples. On average about three-fourths of the cations are Ca and Mg for the Brahmaputra river system. In the Dhansiri, the Kopili, and the Ranganadi tributaries, Ca and Mg account for ⬃60% of the total cation budget. If all the Ca and Mg in all these river waters are derived from carbonate weathering, ⬃75% of the cations (molar) in the Brahmaputra watershed will be of carbonate origin. This places an upper limit on the carbonate weathering contribution of cations, because part of Ca and Mg will also be derived from silicates, and Ca also comes from evaporites. In the Tsangpo tributaries, Ca and Mg account only for 43% of cations (Hu et al., 1982) owing to supply of Na (or K) by dissolution of saline and alkaline salts present in the Tibetan drainage and contribution from saline lakes. Ca/Mg molar ratios in the Brahmaputra river system average ⬃3, with a range of 0.8 to 7.1. The Ca/Mg ratio in rivers would depend on Ca and Mg supplied from various sources—silicates, carbonates, and evaporites—and the behaviour of Ca in the rivers. In the Brahmaputra river system, calculations showed that Ca is undersaturated with respect to calcite in all samples analyzed except in six of them. Even in these six samples, the extent of supersatuartion is marginal, a factor of two or less. Therefore the Ca/Mg variation seems to be more dependent on their sources: Low Ca/Mg can result from silicate/dolomite weathering and high Ca/Mg from calcite weathering. Figure 6 is a plot of Ca/Mg vs. Mg concentration which shows distinct decrease with increase in Mg. This is an indication of the role of dolomite weathering in contributing to Ca and Mg. The regional lithology and detrital carbonate composition of the Dibang, the Lohit, and the Manas rivers attest to this inference. The anion budget is dominated by alkalinity (Fig. 5b), which varies from 425 to 1877 ␮M (Table 2), contributing ⬃90% (molar) to the anion budget. The temporal variation in alkalinity of the Brahmaputra mainstream at Guwahati (Table 3), from 1025 to1694 ␮M, is less than its spatial variation along its course. SO4 is the next abundant anion. Analogous to alkalin- 4.2. Major Ion Composition Downstream variation in the total cation charge in the Brahmaputra mainstream ranges from ⬃2350 ␮Eq L⫺1 in Tibet to 1085 ␮Eq L⫺1 at Tezpur (Table 2). This is quite similar to the temporal variability in total cation (TZ⫹) at Guwahati, 1242 to 2641 ␮Eq L⫺1, and the range in the tributaries, 604 to 2154 ␮Eq L⫺1. The TZ⫹ charge balances total anions (TZ⫺) in most samples within analytical uncertainties. The normalized inorganic charge balance (NICB) is within ⫾5% for most of the samples. In four samples TZ⫺ exceeds TZ⫹ by 5% to 11% with the maximum deviation for the Ranganadi sample. The “excess” anions can be due to the presence of organic ligands such as oxalate, acetate, and humic/fulvic acids. The cation budget (Fig. 5a) is dominated by Ca and Mg in Fig. 6. Scatter diagram of Ca/Mg vs. Mg. Decrease in Ca/Mg is due to increase in Mg, likely due to dissolution of dolomite of the Lesser Himalaya; exposure of such lithologies increases from east to west. Chemical erosion in the eastern Himalaya ity, temporal variation in SO4 of the Brahmaputra mainstream at Guwahati (Table 3, Fig. 4), 85 to 185 ␮M (barring one sample with 456 ␮M SO4; see discussion), is also less than its spatial variation. The SO4 content of the Himalayan tributaries, in general, is lower compared to that of the Ganga and the Yamuna (Sarin et al., 1992; Dalai et al., 2002). This is consistent with the lithology of their basins, which have lesser exposure of the Lesser Himalaya sedimentaries and hence less pyrites and evaporites in their watershed. SO4 content of the Himalayan tributaries of the Brahmaputra in the Assam plain increases from east to west following the increase in the proportions of the Lesser Himalayan sedimentaries. The SO4 content of the Brahmaputra is generally higher than that of the Ganga (Sarin et al., 1992; Galy and France-Lanord, 1999), contributed from Tibetan tributaries. Cl in the Brahmaputra River System is quite low. It ranges from 9 to 83 ␮M (Table 2), with most samples having values centering around 15 ␮M, similar to those reported for the Himalayan glaciers (Nijampurkar et al., 1993; Sarin and Rao, 2002). In general, Cl in the southern tributaries is higher. The highest Cl is in the Tsangpo (Hu et al., 1982; Chen and Guan, 1981), indicating salt dissolution. In the Brahmaputra mainstream, temporal variation in Cl, 18 to 98 ␮M, is quite significant, with higher concentrations during lean flow. Silica in the Brahmaputra river system varies from 105 to 310 ␮M (Table 2). In Tibet, the Tsangpo has ⬃125 ␮M of silica (Hu et al., 1982) which increases to ⬃200 ␮M as the Brahmaputra crosses the Himalaya. The silica in the Himalayan tributaries, 105 to 294 ␮M, is comparable to those in the headwaters of the Ganga and the Yamuna (Sarin et al., 1992; Dalai et al., 2002). The southern tributaries, the Burhi Dihing, the Dhansiri and the Kopili, have higher silica, 208 –310 ␮M, indicating more intense silicate weathering. Silica on average contributes ⬃12% to the TDS. The silica contribution to TDS in the Tsangpo in Tibet is as low as 4%, indicating lower silicate weathering in its drainage in Tibet. The results on temporal variations in the major ion chemistry of the Brahmaputra show that the sample collected in the month of June has elevated concentrations of most of the ions (Table 3, Fig. 5). It has the highest TDS, Ca, and SO4 concentrations and Ca/Mg ratio. The contribution of Si to TDS is quite low (4%) in this sample, similar to that in the Tsangpo (Hu et al., 1982). The major ion chemistry of this sample shows a close resemblance to those of the Tibetan samples. There were reports of a flash flood in the Brahmaputra during the time period when this sample was collected. The flash flood was a result of the bursting of a naturally built dam in Tibet. This dam had restricted water discharge from the Tibetan region into the Brahmaputra. Its break-up enhanced the contribution of water from the Tibetan drainage to the Brahmaputra, which is reflected in the major ion chemistry. Further, damming increases reaction time of water with basin sediments and rocks, elevating the concentrations of major elements in the dissolved phase. Samples collected during the postmonsoon and one monsoon sample were analyzed for ␦13C of DIC. The samples were not poisoned to arrest biologic activity between sampling and analysis; however, no fungus or any deposit was seen during the time of measurement. The ␦13C of the Brahmaputra main channel varies from ⫺14.3‰ to ⫺10.1‰. The ␦13C ranges for the eastern, the Himalayan, and the southern tributaries are 3581 ⫺12.4‰ to ⫺12.6‰, ⫺11.1‰ to ⫺14.3‰, and ⫺13.6‰ to ⫺17.2‰, respectively. These results have been used to assess the role of silicate/carbonate weathering in the budget of DIC in these waters. 4.3. Chemical Weathering in the Brahmaputra Watershed In this study an attempt has been made to quantify the chemical erosion rate in this drainage, based on major ion chemistry of rivers waters. In addition, the measured mineralogic and clay composition of sediments has been used to provide independent assessment of the chemical weathering in the basin. 4.3.1. Clay composition and CIA of Sediments The clay content of Brahmaputra mainstream, from Dibrugarh to Guwahati, is only 1.7 to 3 wt% of total sediments (Table 5). The Himalayan tributaries (the Subansiri, the Ranganadi, and the Manas) of the Brahmaputra have similar clay abundance, 1.4 to 3.9 wt%. In contrast, the southern tributary, Kopili, has ⬃21 wt% clay. A direct interpretation of the clay content of these river sediments is that in the Brahmaputra mainstream and in the Himalayan drainage the sediments are poorly weathered, whereas for the sediment of the southern tributary the chemical weathering is relatively more intense. An important uncertainty, however, in using the clay mineral abundance to assess the degree of chemical weathering is size sorting in river sediments, which can significantly alter the clay abundance. In drainage basins subject to high-energy flow, such as in the Brahmaputra, clay minerals and fine particles can be transported out of sediments. Therefore, in addition to abundance of clays, other proxies such as composition of clays have to be used to assess the intensity of chemical weathering. In the Brahmaputra mainstream most of the clay is vermiculite, a mineral resulting from less intense chemical weathering (Millot, 1970). Thus, the clay content of the Brahmaputra sediments and the abundance of vermiculite in them is an indication which shows that these sediments are poorly weathered. This can be attributed to rapid transport of sediments along the course of the river; making this drainage weathering limited (Stallard, 1995). In the southern tributaries, where the clay content is higher and is made of montmorillonite, illite, and chlorite, the chemical weathering is high, a conclusion also attested by the higher concentrations of Si and Na* in their waters. The chemical weathering in a basin can also be gauged from the change in mineralogy and chemical composition of sediments along the river course. Sediment composition can be modified by a number of processes occuring during erosion, transport, recycling, and diagenesis. The mineralogy of sediments of the Brahmaputra River System (Garzanti et al., 2004) show that (1) there is a marginal decrease in the plagioclase/ feldspars (P/F) ratio from the mountain streams to the Assam plains, (2) there is no indication of selective dissolution of plagioclase over the more resistive K-feldspar, (3) clinopyroxenes and amphiboles show similar extent of alteration, and (4) quartz/feldspar (Q/F) ratio, P/F ratio, and hornblende-dominated dense-mineral assemblages remain constant. All these 3582 S. K. Singh, M. M. Sarin, and C. France-Lanord observations infer minimal chemical weathering of the Brahmaputra sediments. In contrast the sediments of the southern tributaries contain pitted and embayed quartz grains, have low P/F ratio, and have abundant microcline, etched clinopyroxene, and laterite clasts. These observations support a more intense weathering in this drainage. This can be attributed to higher temperature in the region and higher residence time of sediments in the basin because of lower relief. The chemical weathering of sediments can also be assessed using their chemical index of alteration (CIA; Nesbitt and Young, 1982): CIA ⫽ 100 ⫻ 兵(Al2O3 ⁄ 共Al2O3 ⫹ CaO * ⫹ Na2O ⫹ K2O)其 where Al2O3, CaO*, Na2O, and K2O are molar abundances. CaO* is CaO corrected for carbonates present in the sediments (Nesbitt and Young, 1982). CIA of fresh granites is ⬃50, which on weathering increases to ⬃100 in soils. CIA for rocks from the Higher and Lesser Himalaya are ⬃65–70 as these are recycled crust whereas those for the Transhimalayan Plutonic rocks are ⬃55 (calculated from Debon et al., 1986), being the recent calc-alkaline rocks of mantle origin. Bedloads of the Brahmaputra mainstream have CIA ranging from 58 to 64 (data from Singh and France-Lanord, 2002). (These CIA calculations are based on total Ca, as these sediments have negligible amount of carbonate, Singh and France-Lanord, 2002. In the Siang, the Lohit, the Dibang, and the Manas sediments, which have carbonate content of ⬃3%– 4%, CaO is appropriately corrected for carbonates to calculate CIA. One gneiss sample collected from Guwahati has CIA ⬃58. These lower CIA values indicate that they are less weathered. The lower CIA of the Brahmaputra can arise due to the presence of Transhimalayan Plutonic rocks in the basin of the Siang and the eastern tributaries which have CIA values of ⬃55 (from data Debon et al., 1986). The eastern and the Himalayan tributaries have CIA values 56 –59 and 65–75, respectively, which underscores the importance of Transhimalayan Plutonic rocks in lowering the CIA of the Brahmaputra. The CIA of the southern tributaries range from 65 to 82, indicating that sediments of these tributaries are more weathered. This is also consistent with the inferences based on mineralogy and clay content of sediments and dissolved major ion composition of corresponding rivers. 4.3.2. Sources of Solutes in the Brahmaputra River System Determining the sources of solutes, particularly silicate derived solutes, in the Brahmaputra river system is one of the major objectives of this study as it has relevance to CO2 consumption and hence to global change. Rivers receive solutes from several sources, precipitation and weathering of silicate and carbonate rocks and dissolution of evaporites. To determine the silicate and carbonate contributions to the major ion chemistry, inputs from rain and evaporites have to be properly accounted for. This is commonly achieved using Cl as index (Sarin et al., 1989; Dalai et al., 2002; Krishnaswami et al., 1999; Singh et al., 1998) and assuming Na ⫽ Cl. The Na corrected for Cl (Na* ⫽ Nariv ⫺ Clriv) is generally taken to be of silicate origin. In the case of the Tsangpo, as mentioned earlier, however, there can be other sources of Na such as sodium carbonate or borax in the evaporite deposits (Pascoe, 1963). In such cases, correction only for evaporites using Cl as Fig. 7. Si vs. Na*. Good correlation between Si and Na* (Nariv ⫺ Clriv) indicates that Na* can be used as a proxy of silicate weathering. Si/Na* for most of the rivers is 2.0, indicating plagioclase weathering to kaolinite, whereas for the Eastern tributaries it is ⬃4.0, owing to weathering of mafic/ultramafic rocks in their drainage. Rivers influenced by evaporite dissolution such as the Tsangpo, its tributaries in Tibet, and Dhansiri of the southern drainage fall off the line, indicating contribution of Na from evaporites. an index can yield an upper limit of Na from silicates. Rainwater contribution to other cations and anions are insignificant (e.g., Galy and France-Lanord, 1999) and therefore no correction is applied. Figure 7 presents a covariation between Na* and Si in the rivers analysed. All data, except five, fall in a line, showing good correlation (r2 ⫽ 0.86), attesting to the idea that Na* in the rivers can serve as a silicate weathering index. The five data points which fall off the general trend and have “excess” Na* are for the Tsangpo, Tibetan tributaries, and one of the southern tributaries; in these rivers there could be additional sources for Na, such as evaporite minerals. Si/Na* for most of the rivers is ⬃2.0 which is similar to the slope of the line, 1.84 (Fig. 7), indicating plagioclase weathering to kaolinite. The eastern tributaries, the Dibang, and the Lohit have Si/Na* ⬃4. High Si/Na* in these rivers is possibly due to weathering of sodiumdeficient pyroxenes. These inferences, in general, are supported by the mineralogy of sediments. It is assumed that all K is of silicate origin because evaporites and carbonates have very little K. Determination of silicate-derived Ca and Mg is prone to more uncertainties because these elements are supplied to rivers from multiple sources: from silicates, carbonates, and evaporites. Even among silicates, Ca and Mg contributions can be significantly different depending on their composition and weatherability. Therefore, to obtain silicate Ca and Mg contribution it is necessary to know the composition of silicate rocks of the basins. Singh and France-Lanord (2002) have shown, based on Sr and Nd isotopes, that the Brahmaputra sediments are a mixture of the Transhimalayan Plutonic Belt rocks and those from the Himalaya in the proportion 30:70. Using the Ca, Mg, and Na abundances of these endmembers and above mixing proportion, the Ca/Na and Mg/Na ratios for the silicates of the Brahmaputra watershed were calculated. The elemental Chemical erosion in the eastern Himalaya 3583 Table 6. Endmember of silicates for various drainages. River Lohit, Dibang Ranganadi, Jia Bhareli, Manas, Puthimari, Burhi Dihing, Dhansiri, Kopili, Tipkai Tsangpo, Siang, Subansiri, Brahmaputra mainstream a b Ca/Na (molar) Mg/Na (molar) (Ca ⫹ Mg)/Na (molar) Transhimalaya Himalaya 2.7 ⫾ 1.4a 0.7 ⫾ 0.4b 2.5 ⫾ 1.3a 0.3 ⫾ 0.2b 5.2 ⫾ 2.6 1.0 ⫾ 0.5 Transhimalaya ⫹ Himalaya (30:70) 1.3 ⫾ 0.7 1.0 ⫾ 0.5 2.25 ⫾ 1.1 Formation Debon et al. 1986. Krishnaswami et al., 1999. ratios of various endmembers and the rivers which drain them are given in the Table 6. The endmember composition have been assigned an uncertainty of ⫾50% owing to variation in their mixing proportions and the approaches adopted for calculation (Krishnaswami et al., 1999). The Ca⫹Mg derived from the silicates, (Ca⫹Mg)sil, is calculated using these endmember ratios (Table 6) and Na* of the river waters (Table 2) using the relation (Singh et al., 1998) 共Ca ⫹ Mg兲sil ⫽ 关共Ca ⫹ Mg兲 ⁄ Na兴 ⫻ Na* The results show that (Ca⫹Mg)sil in these waters varies from 9% to 59% of total Ca⫹Mg (molar), with a mean of 31%. The total cations derived from silicate, (⌺Cat)sil, is calculated by summing Na*, K, and (Ca⫹Mg)sil. These vary between 21% and 77% in the rivers of the Brahmaputra River System, with a mean of 45%, nearly the same as (⌺Cat)sil of the Brahmaputra mainstream at Chilmari, its outflow. The (⌺Cat)sil will be ⬃45% ⫾ 11% if the 50% uncertainties in the (Ca⫹Mg)/Na ratio of silicate endmember (Table 6) is considered. The (⌺Cat)sil generally follows the lithology of the individual drainage. The major uncertainty in the (⌺Cat)sil is from Ca/Na of the silicate endmember. Galy and France-Lanord (1999) assumed a value of 0.2 for Ca/Na of Himalayan silicate, whereas in this study it has been taken as 0.7. If the value of 0.2 is used for Ca/Na for the Himalaya, the (⌺Cat)sil will vary between 16% and 60%, with a mean of 38%. While the value of 0.2 is based on the Ca/Na of silicates from HH, LH, and leucogranites (Galy and France-Lanord, 1999), a value of 0.7 is derived for Ca/Na based on the Ca/Na of silicates (crystallines and sedimentaries) of LH and HH, of the soil profiles developed over LH silicates, and of stream flowing exclusively through silicate terrains (Krishnaswami et al., 1999). Calculating (⌺Cat)sil in the Tsangpo (Tibet) following the above approach has been hampered, because the validity of using Cl as an index for rain and evaporite contribution of Na is in doubt. In this drainage, in addition to rain and halites, Na may also be derived from sulphates, carbonates, and borates (Pascoe, 1963). Therefore in this drainage silicate cations have been calculated based on the overall relation (Fig. 7) between Si and Na*. Na* from silicates is estimated from the Na-Si relation (Fig. 7) and the SiO2 concentration of ⬃125 ␮M for the Tsangpo. The calculated Na* is ⬃45 ␮M. Based on this Na*, a (⌺Cat)sil of 150 ⫾ 25 ␮M has been estimated for the Tsangpo; this represents ⬃17% of the total cations, indicating silicate cation contribution to total cation is quite low. This, as mentioned earlier, is because of supply of higher proportion of alkalis from alkaline and saline salts of the basin. 4.3.3. Erosion in Various Subbasins Singh and France-Lanord (2002), based on Sr and Nd isotope composition of sediments from the Brahmaputra drainage, inferred that this drainage undergoes differential physical erosion. As a part of the current study, chemical erosion in the various subbasins of the Brahmaputra watershed have been determined to quantify their total chemical and silicate erosion rates and to evaluate the factors influencing chemical erosion. Erosion rates of various subbasins of the Brahmaputra River System are given in Table 7 based on the TDS and (⌺Cat)sil fluxes. For comparison, the erosion rates of a few other selected basins of the Himalaya and of other global basins (Sarin et al., 1989) are also given in this table. For the Eastern Syntaxis zone of the Siang, TDS and (⌺Cat)sil fluxes are taken to be the difference in TDS and (⌺Cat)sil fluxes from Tibet at Pai and those at Pasighat. Table 7 shows that chemical erosion in the Brahmaputra watershed is not uniform and that it varies by more than an order of magnitude among the various subbasins. Both total chemical erosion and silicate erosion rates in the Eastern Syntaxis zone are the highest compared to other sub- Table 7. Chemical erosion and CO2 consumption rates in various sub-basins of the Brahmaputra and selected basins of the worlda Basin Brahmaputra Tibet Eastern Syntaxis Eastern Himalaya Southern Brahmaputra Other Himalayan Rivers Ganga Indus Mekong Global Rivers Amazon World Average a Silicate Cation Flux CO2 consumption by silicate weathering, 105 moles km⫺2 y⫺1 40 304 185 149 237 120 1.3 38.0 18.1 10.7 22 11.8 0.7 19 9.5 5.1 12 6 72 42 72 7.9 1.8 6.2 3.8 0.6 2.4 35 36 2.2 2.0 0.5 0.9 TDS Flux (t km⫺2 y⫺1) Data for TDS of other rivers are from Sarin et al. (1989) and references therein, and those for Silicate Cation fluxes and CO2 consumption due to silicate weathering are from Gaillardet et al. (1999) and Krishnaswami et al. (1999). 3584 S. K. Singh, M. M. Sarin, and C. France-Lanord basins. The chemical erosion rate in this zone is even higher than those reported for some of the basaltic terrains such as the Reunion (Louvat and Allegre, 1997) and Iceland (Gislason et al., 1996). This zone is also characterized by very high physical erosion, contributing about half of the sediments to the Brahmaputra at its outflow (Singh and France-Lanord, 2002). The southern tributaries are also eroding quite rapidly but their total impact on the Brahmaputra River System is not significant, contributing only ⬃8% of the discharge. The higher chemical erosion in the Eastern Syntaxis zone is attributable to the higher physical erosion in this region brought about by higher runoff and steep gradient of the Siang River. This inference attests to the dependence of chemical erosion on runoff probably through a corresponding enhancement in physical erosion (Bluth and Kump, 1994). The runoff as well as gradient of the Tsangpo are low in Tibet, leading to lower physical and hence correspondingly lower chemical and silicate erosion rates. 4.3.4. ␦13C of Dissolved Inorganic Carbon (DIC) A major uncertainty in calculating the silicate contribution of major ions to rivers is associated with (Ca/Na) and (Mg/Na) ratios supplied to dissolved phase from various silicate endmembers. Therefore, to determine silicate weathering contribution to major ions of rivers, the potential of ␦13C has been exploited (Galy and France-Lanord, 1999; Karim and Veizer, 2000). Sources of alkalinity in river waters were quantified using ␦13C of DIC. ␦13C of DIC depends on the source of carbon, which are CO2 from atmosphere and soil and from carbonates released during their weathering. The contribution of alkalinity from solution of atmospheric CO2 should be negligible in the Brahmaputra waters, which have ⱖ300 ␮M 13 HCO⫺ of soil CO2 in the Brahmaputra basin is about 3 . The ␦ ⫺26‰ because of the dominance of C3 plants in this region (Burbank et al., 1993). This soil CO2 during weathering of silicates will produce alkalinity with ␦13C ⬃⫺18% (Solomon and Cerling, 1987) and with ⬃⫺9‰ in carbonate weathering, because half of the alkalinity during carbonate weathering is from carbonate and the other half from soil CO2. The Burhi Dihing has the most depleted ␦13C, indicating highest silicatederived alkalinity. This is also consistent with its low Ca/Na* showing higher contribution of silicate cations. In contrast, the Manas has more enriched ␦13C and high Ca/Na*, suggesting higher contribution from carbonates. Based on calculated endmember ␦13C values for silicate- and carbonate-derived DIC and the measured ␦13C of DIC in the Brahmaputra River System (⫺10.1‰ to ⫺17.2‰, Table 2), it can be estimated that on average about half of the alkalinity in these waters is derived from silicate weathering and the balance is supplied by carbonate weathering. The silicate alkalinity estimated using ␦13C of river water roughly balances the silicate cations calculated based on the major ion data in earlier section. 4.3.5. CO2 consumption by silicate weathering One of the objectives of this study is to calculate CO2 consumption due to silicate weathering, because this is a key parameter to understand the coupling between chemical weathering in the Himalaya and drawdown of atmospheric CO2 (Raymo and Ruddiman, 1992). The rates of CO2 consumption by silicate weathering in the Brahmaputra drainage and for its various subbasins have been calculated based on the cations derived from silicates and assuming that all weathering results only from CO2. The CO2 consumption by silicate weathering for the entire Brahmaputra basin is calculated to be ⬃6 ⫻ 105 moles km⫺2 y⫺1 based on the monsoon sample from Dhubri. This is comparable to the estimate based on Si (1 mol Si ⫽ 2 mol of CO2; Huh et al., 1998) of ⬃4.5 ⫻ 105 moles km⫺2 y⫺1. Further, the discharge-weighted CO2 consumption at Guwahati calculated from the time series data of chemical composition is ⬃7 ⫻ 105 moles km⫺2 y⫺1, similar to estimate of ⬃6.4 ⫻ 105 moles km⫺2 y⫺1 based on ␦13C of DIC and marginally higher than the value of 5.3 ⫻ 105 moles km⫺2 y⫺1 for the monsoon sample and (Table 2). These values of CO2 consumption for the Brahmaputra, though, seem higher than those obtained for the Alaknanda, Bhagirathi, and the G-B (Krishnaswami et al., 1999); they fall within the range for the rivers in the Himalaya, 2–7 moles km⫺2 y⫺1 (Dalai et al., 2002). Further, the CO2 consumption due to silicate weathering in the Himalayan drainage of the Brahmaputra is closer to those reported for the Alaknanda and the Bhagirathi (Krishnaswami et al., 1999). The value ⬃6 ⫻ 105 moles km⫺2 y⫺1, in this study may be an upper limit as part of silicates may also be weathered by the H2SO4 produced from pyrite oxidation. However the observation that HCO3/SO4 (equivalent) averages ⬃6, and that Ca/SO4 is high, ⬃4, it seems that weathering mediated by H2SO4 may not be significant in this basin. Negligible contribution of silicate alkalinity to these rivers due to weathering by H2SO4 is also supported by the CO2 consumption rate estimate of ⬃6.4 ⫻ 105 moles km⫺2 y⫺1 based on ␦13C of DIC at Guwahati which is similar to the estimate by major ion. The annual CO2 consumption by silicate weathering in the Brahmaputra basin is ⬃2.72 ⫻ 1011 moles y⫺1 at Dhubri. This estimate is comparable to those reported for the Brahmaputra using Si content by Galy and France-Lanord (1999), however, it is higher than their estimate based on major ions due to difference in endmember estimates of Ca/Na. The silicate weathering in the Brahmaputra River System consumes ⬃ 2%–3% of the global consumption of CO2 by silicate weathering (Amiotte Suchet et al., 2003; Gaillardet et al., 1999) with only ⬃1.6% of the global water discharge. Chemical erosion among the different subbasins of the Brahmaputra is highly variable and therefore their CO2 consumption rates also are expected to show similar variability. The CO2 consumption for subbasins (Table 7) vary between 0.7 and 19 ⫻ 105 moles km⫺2 y⫺1, with the Eastern Syntaxis basin having the highest consumption, 19 ⫻ 105 moles km⫺2 y⫺1. This rate for the eastern syntaxis is much higher than the Brahmaputra average but is similar to the CO2 consumption rates reported for the basaltic terrains such as the Reunion (Louvat and Allegre, 1997) and the Iceland (Gislason et al., 1996; Fig. 8). This is an important result and suggests that granite/gneisses and mafic rocks of the Brahmaputra river basin, can also erode as rapidly as basalts from other regions under “favorable” conditions. A similar observation is also made by Das et al. (2005). The lowest CO2 consumption by silicate weathering among the subbasins of the Brahmaputra is for the Tibetan subbasin, attributable to factors such as its flat topography, lower runoff, and colder climate. Chemical erosion in the eastern Himalaya 3585 Fig. 8. CO2 consumption rate due to silicate weathering for various subbasins and for the entire Brahmaputra basin. Values for the Ganga, the Yamuna, Reunion, and Iceland are taken from Dalai et al. (2002), Louvat and Allegre (1997), and Gislason et al. (1996), respectively. The rate of CO2 consumption due to silicate weathering for the Eastern Syntaxis region is notably high and comparable to the basaltic terrain such as those for Reunion and Iceland. Higher CO2 consumption in this river section is due to higher physical weathering caused by higher runoff and stream gradient (stream power) of this zone. 4.4. Controls on Chemical Erosion in the Brahmaputra Drainage It has been suggested that several factors control chemical erosion. These include temperature, physical erosion, run off, lithology, elevation, vegetation ,and tectonic activities (Velbel, 1993; White and Blum, 1995; Berner and Berner, 1997; Edmond and Huh, 1997; Huh and Edmond, 1999; Millot et al., 2002; France-Lanord et al., 2003; Dalai et al., 2002). 4.4.1. Runoff vs. Chemical Erosion Runoff has been suggested as a major factor controlling chemical erosion, (Bluth and Kump, 1994; France-Lanord, 2003; Gaillardet et al., 1999). The eastern Himalaya receives heavy rainfall during both SW and NE monsoons, and therefore runoff in this region varies both in space and time. This makes the Brahmaputra drainage a suitable location to assess the impact of runoff on chemical weathering. The TDS flux (t km⫺2 y⫺1), an index of chemical erosion, has been plotted against runoff in Figure 9a. TDS data during SW monsoon has been considered for this plot because most of the annual river discharge occurs in this period. The results show a good positive correlation between runoff and the TDS flux (r2 ⫽ 0.81), attesting to the earlier suggestions that runoff exerts significant control over chemical erosion. A similar trend is also observed if (⌺Cat)sil flux is plotted against runoff (Fig. 9b). In the Brahmaputra basin, higher runoff causes higher physical erosion exposing fresh surfaces for chemical erosion. Thus, runoff and physical ero- Fig. 9. TDS (a) and (⌺Cat)sil (b) fluxes vs. runoff. The strong correlation between TDS flux, (⌺Cat)sil flux and runoff indicate that runoff is major controlling factor of the chemical erosion. sion significantly influence chemical erosion of the Brahmaputra basin. It can be argued that the above inference is not unexpected, because runoff is a multiplication factor for calculating chemical erosion rates (CER ⫽ TDS ⫻ runoff). In samples plotted in Figure 9a, the runoff varies by more than an order of magnitude among the various river basins whereas the TDS abundance variation is only within a factor of ⬃2, indicating the importance of runoff in controlling CER. 4.4.2. Physical vs. Chemical Erosion The role of physical erosion as a “driver” of chemical erosion has been established in many river basins (Bluth and Kump, 1994; Millot et al., 2002). Physical erosion is quite high in the rivers of the Himalaya and in the Brahmaputra in particular. Figure 10 is a scatter plot of physical erosion rates of the various subbasins of the Brahmaputra vs. their chemical erosion rates. Physical erosion rates (PER) have been calculated based on the proportion of sediment supplied by these 3586 S. K. Singh, M. M. Sarin, and C. France-Lanord Table 8. Annual fluxes of cations and TDS from the Brahmaputra. Flux Cations and TDS a b Na K Mg Ca TDS 45 41 77 308 73 116 107 202 803 189 a Na, K, Mg, and Ca in 109 mol y⫺1; TDS in 106 t y⫺1. b Na, K, Mg, and Ca in 103 mol km⫺2 y⫺1; TDS in t km⫺2 y⫺1. Fig. 10. Physical vs. chemical erosion. A significant coupling between chemical and physical erosion is observed, which is related by a power law. The chemical erosion in the Brahmaputra watershed is controlled by the physical erosion, which is primarily controlled by runoff and relief. basins (Singh and France-Lanord, 2002) and a sediment discharge of 700 million t y⫺1 for the Brahmaputra. Sediment flux from the southern subbasin is not significant and therefore not plotted. Figure 10 shows a strong coupling between physical and chemical erosion, with the following power law: CER ⫽ 4.61(PER)0.44 Millot et al. (2002) also report a power law relation between the chemical and physical erosion rates for a few granitic and basaltic river basins; however, values of the constant and exponent of their relation are different from those derived above for the Brahmaputra. Comparison of these relations indicates that for a given PER the CER will be more in the case of the Brahmaputra compared to other river systems (Millot et al., 2002). In the lower PER range this difference is particularly significant, possibly owing to the higher temperature of the tropical region. From the above discussion and positive correlation between runoff and PER (Singh and France-Lanord, 2002) it can be inferred that higher runoff causes higher physical erosion which provides more fresh material for chemical erosion. Therefore runoff exerts the first-order control on the chemical erosion. Further, the data show that enhanced runoff and higher relief are the primary variables that promote physical erosion, which in turn increases the chemical erosion. 4.5. Chemical Fluxes from the Brahmaputra The bimonthly chemical compositions of the Brahmaputra at Guwahati coupled with monthly discharge data for Pandu (near Guwahati, GRDC website) have been used to calculate the annual chemical fluxes from the Brahmaputra at Guwahati. The calculated fluxes of various cations and TDS are given in Table 8. Chemical composition of the water of the Brahmaputra at Guwahati are not measured for August and September. To calculate the discharge-weighted annual fluxes of cations and TDS, July concentrations are assumed for the waters of August and September. Similar to July, August and September fall in SW monsoon season; therefore this assumption is expected to be valid. The estimated fluxes are higher for K, Ca, and TDS and lower for Na and Mg than those reported by Sarin et al. (1989) and Galy and France-Lanord (1999). In our study fluxes have been calculated at Guwahati, whereas in Sarin et al., (1989) and in Galy and France-Lanord (1999) it is at the outflow. The difference arises from the temporal variability observed in the concentration of cations, TDS, and discharge and reinforces the need for temporal data of the river water chemistry and its discharge for calculating fluxes. Based on these fluxes, it can be concluded that the Brahmaputra drainage is eroding fast chemically, ⬃2 to 3 times faster than the Ganga drainage (Sarin et al., 1989; Galy and France-Lanord, 1999) and ⬃5 times higher than the global average (Sarin et al., 1989). 5. CONCLUSIONS Chemical erosion in the Brahmaputra drainage transports ⬃73 million tons of dissolved material annually to the world oceans; representing ⬃4% of the total dissolved load via world rivers. About half of the dissolved cations are of silicate origin, which is higher than that reported for the Ganga basin. Such characteristic difference could arise owing to diverse lithologies of the two basins: Ganga basin is relatively dominated by sedimentaries (containing more easily weatherable carbonates), whereas the Transhimalayan mafic/ultramafic rocks occur in the Brahmaputra drainage. The chemical erosion rate in the Brahmaputra (mainly governed by runoff and physical erosion) is ⬃4 –5 times higher than the global average. However, there exists a significant spatial variability in the chemical erosion; some of the regions are eroding relatively faster, particularly the region around the Eastern Syntaxis. Despite the fact that higher chemical erosion rates are representative in the Brahmaputra basin compared to that in the Ganga, the intensity of chemical weathering, is relatively lower in the Brahmaputra basin. The cause for this is attributed to enhanced physical erosion and rapid sediment transport in the Brahmaputra drainage. CO2 consumption due to silicate weathering in the Brahmaputra drainage is ⬃6.0 ⫻ 105 moles km⫺2 y⫺1, which is a few times higher than the global average. The Chemical erosion in the eastern Himalaya Brahmaputra drainage accounts for ⬃2%–3% of global CO2 consumption due to silicate weathering, marginally higher than its contribution to water discharge ⬃1.6%. Analogous to chemical and silicate erosion rates, higher CO2 consumption rate in the Eastern Syntaxis zone is comparable to some of the global basaltic terrains. Acknowledgments—We thank S. Krishnaswami for the useful discussion throughout the course of this study. The analytical help provided by A. K. Sudheer was extremely valuable. We are greatly benefited from the comments provided by the three anonymous reviewers. The evaluation report provided by Lee Kump was helpful. Associate editor: Lee Kump REFERENCES Amiotte Suchet P., Probst J-L., and Ludwig W. (2003) Worldwide distribution of continental rock lithology: Implications for the atmospheric/soil CO2 uptake by continental weathering and alkalinity river transport to the oceans Global Biogeochem. Cycles 17, 1038. Bartoli F., Burtin G., and Herbillon A. J. (1991) Desegregation and clay dispersion of oxisols: Na resin, a recommended methodology. Geoderma 49, 301–317. Berner E. K. and Berner R. A. (1997) Silicate weathering and climate. In Tectonic Uplift and Climate Change (ed. W. F. Ruddiman). Plenum, New York, pp.354 –365. Bickle M. J., Bunbury J., Chapman H. J., Harris N. B. W., Fairchild I. J., and Ahmad T (2003) Fluxes of Sr into the headwaters of the Ganges. Geochim. Cosmochim. Acta 67, 2567–2584. Blum J. D., Gazis C. A., Jacobson A. D., and Chamberlain C. P. (1998) Carbonate versus silicate weathering in Raikhot watershed within the High Himalayan crystalline series. Geology 26, 411– 414. Bluth G. J. S. and Kump L. R. (1994) Lithologic and climatologic controls of river chemistry. Geochim. Cosmochim. Acta 58, 2341– 2359. Burbank D. W., Derry L. A., and France-Lanord C. (1993) Reduced Himalayan sediment production 8 Myr ago despite an intensified monsoon. Nature 364, 48 –50. Burg J.-P., Nievergelt P., Oberli F., Seward D., Davy P., Maurin J.-C., Diao Z., and Meier M. (1998) The Namche Barwa syntaxis: evidence for exumation related to compressional crustal folding. J. Asian Earth Sci. 16, 239 –252. Chen C., and Guan Z. (1981) Hydrochemistry of rivers in Xizang. In Geological and Ecological Studies of Qinghai-Xizang Plateau. Gordon and Brach Science Publishers, New York, pp. 1687–1692. Dalai T. K., Krishnaswami S., and Sarin M. M. (2002) Major ion chemistry in the headwaters of the Yamuna river system: Chemical weathering, its temperature dependence and CO2 consumption in the Himalaya. Geochim. Cosmochim. Acta 66, 3397–3416. Das A., Krishnaswami S., Sarin M. M., and Pande K. (2005) Chemical weathering in the Krishna basin and Western Ghats of the Deccan Traps, India: Rates of basalt weathering and their controls. Geochim. Cosmochim. Acta 69, 2067–2084. Debon F., Le Fort P., Sheppard S. M. F., and Sonet J. (1986) The four plutonic belts of the Transhimalaya-Himalaya: a chemical, mineralogical, isotopic and chronological synthesis along a Tibet-Nepal section. J. Petrol. 27, 219 –250. Edmond J. M. and Huh Y. (1997) Chemical weathering yields and orogenic terrains in hot and cold climates. In Tectonic Uplift and Climate Change (ed. W. F. Ruddiman). Plenum, New York, pp. 330 –351. France-Lanord C., Evans M., Hurtrez J. -E., and Riotte J. (2003) Annual dissolved fluxes from Central Nepal rivers: Budget of chemical erosion in the Himalayas. Comptes Rendus Geosci. 335, 1131–1140. Gaillardet J., Dupre B., and Allegre C. J. (1999) Global silicate weathering and CO2 consumption rates deduced from chemistry of large rivers. Chem. Geol. 159, 3–30. 3587 Galy A. and France-Lanord C. (2001) Higher erosion rates in the Himalaya: geochemical constraints on riverine fluxes. Geology 29, 23–26. Galy A. and France-Lanord C. (1999) Weathering processes in the Ganges-Brahmaputra basin and the riverine alkalinity budget. Chem. Geol. 159, 31– 60. Galy A., France-Lanord C., and Derry L. A. (1999) The strontium isotopic budget of Himalayan rivers in Nepal and Bangladesh. Geochim. Cosmochim. Acta 63, 1905–1925. Gansser A. (1964) Geology of the Himalaya. Interscience Publishers, Londonp. 289. Garzanti E., Vezzoli G., Andò S., France-Lanord C., Singh S. K., and Foster G. (2004) Sand petrology and focused erosion in collision orogens:The Brahmaputra. Earth Planet. Sci. Lett. 220, 157–174. Gislason S. R., Amorsson S., and Armannsson H. (1996) Chemical weathering of basalt as deduced from the composition of precipitation, rivers and rocks in SW Iceland. Am J. Sci. 296, 837–907. Goswami D. C. (1985) Brahmaputra River, Assam, India: Physiography, basin denudation and channel aggradation. Water Resources Res. 21, 959 –978. Guan Z. and Chen C. (1981) Hydrographical features of the Yarlung Zangbo River. In Geological and Ecological studies of QinghaiXizang Plateau. Gordon and Brach Science Publishers, New York, pp. 1693–1703. Harris N., Bickle M., Chapman H., Fairchild I., and Bunbury J. (1998) The significance of the Himalayan rivers for silicate weathering rates: Evidence from the Bhote Kosi tributary. Chem. Geol. 144, 205–220. Hay W. W. (1998) Detrital sediment fluxes from continents to oceans. Chem. Geol. 145, 287–323. Hu M., Stallard R. F., and Edmond J. (1982) Major ion chemistry of some large Chinese Rivers. Nature 298, 550 –553. Huh Y. and Edmond J. (1999) The fluvial geochemistry of rivers of Eastern Siberia: III. Tributaries of the Lena and Anbar draining the basement terrain of the Siberian Craton and the Trans-Baikal Highlands. Geochim. Cosmochim. Acta 63, 967–987. Huh Y, Tsoi M.-Y., Zaitsev A., and Edmond J. (1998) The fluvial geochemistry of the rivers of Eastern Siberia: I. Tributaries of the Lena River draining the sedimentary platform of the Siberian Craton. Geochim. Cosmochim. Acta 62, 1657–1676. Jain A. K. and Thakur V. C. (1978) Abor volcanics of the Arunachal Himalaya. J. Geol. Soc. India. 19, 335–349. Karim A. and Veizer J. (2000) Weathering processes in the Indus River Basin: Implications from riverine carbon, sulfur, oxygen and strontium isotopes. Chem. Geol. 170, 153–177. Krishnaswami S., Trivedi J. R., Sarin M. M., Ramesh R., and Sharma K. K. (1992) Strontium isotopes and rubidium in the GangaBrahmaputra river system: Weathering in the Himalaya, fluxes to the Bay of Bengal and contribitions to the evolution of oceanic 87 Sr/86Sr. Earth Planet. Sci. Lett. 109, 243–253. Krishnaswami S., Singh S. K. and Dalai T. (1999) Silicate weathering in the Himalaya: Role in contributing to major ions and radiogenic Sr to the Bay of Bengal. In Ocean Science, Trends and Future Directions (ed. B. L. K. Somalyajulu). Indian National Science Academy and Akademia International, New Delhi, pp. 23–51. Kumar G. (1997) Geology of the Arunachal Pradesh. Geological Society of India. Louvat P. and Allegre C. J. (1997) Present denudation rates on the island of Reunion determined by river chemistry: Basalt weathering and mass budget between chemical and mechanical erosions. Geochim. Cosmochim. Acta 61, 3645–3699. Ludwig W. and Probst, J.-L. (1998) River sediment discharge to the oceans: Present-day controls and global budgets. Am. J. Sci. 298, 265–295. Milliman J. D. and Meade R. H. (1983) World delivery of river sediment to the oceans. J. Geol. 1, 1–21. Milliman J. D. and Syvitski P. M. (1992) Geomorphic/tectonic control of sediment discharge to the ocean: The importance of small mountainous rivers. J. Geol. 100, 525–544. Millot G. (1970) Geology of Clays. Springer-Verlag, New York, p. 430. Millot R., Gaillardet J., Dupré B., and Allègre C. J. (2002) The global control of silicate weathering rates and the coupling with physical 3588 S. K. Singh, M. M. Sarin, and C. France-Lanord erosion: new insights from rivers of the Canadian Shield. Earth Planet. Sci. Lett. 196, 83–98. Nakata T. (1989) Active faults of the Himalaya of India and Nepal. In Tectonics of the Western Himalayas (eds. L. L. Malinconico, and R. J. Lillie).Geol. Soc. Am. Spec. Pap. 232, 243–264. Nesbitt H. W. and Young G. M. (1982) Early proterozoic climate and plate motion inferred from major element chemistry of lutites. Nature 299, 715–717. Nijampurkar V. N., Sarin M. M., and Rao D. K. (1993) Chemical composition of snow and ice from Chhota Shigri glacier, Central Himalaya. Jour. Hydrol. 151, 19 –34. Palmer M. R., and Edmond J. M. (1989) The strontium isotope budget of the modern ocean. Earth Planet. Sci. Lett. 92, 11–26. Pascoe, E. H. (1963) A Manual of the Geology of India and Burma Vol. 3. Government of India Press, Calcutta, pp. 2073–2079. Rao K. L. (1979) India’s Water Wealth. Orient Longman, New Delhi, p. 267. Raymo M. E. and Ruddiman W. F. (1992) Tectonic forcing of late Cenozoic climate. Nature 359, 117–122. Richter F. M., Rowley D. B., and DePaolo D. J. (1992) Sr isotope evolution of seawater: The role of tectonics. Earth Planet. Sci. Lett. 109, 11–23. Robinson D. M., DeCelles P. G., Patchett P. J., and Garzione C. N. (2001) The kinematic evolution of the Nepalese Himalaya interpreted from Nd isotopes. Earth Planet. Sci. Lett. 192, 507–521. Sarin M. M., Krishnaswami S., Dilli K., Somayajulu B. L. K., and Moore W. S. (1989) Major ion chemistry of the Ganga-Brahmaputra river system: Weathering processes and fluxes to the Bay of Bengal. Geochim. Cosmochim. Acta 53, 997–1009. Sarin M. M., Krishnaswami S., Trivedi J. R., and Sharma K. K. (1992) Major ion chemistry of the Ganga source waters: Weathering in the high altitude Himalaya. Proc. Indian Acad. Sci. (Earth Planet. Sci.) 101, 89 –98. Sarin M. M. and Rao D. K. (2002) Atmospheric deposition of chemical constituents on a Central Himalayan Glacier: Inference from snow chemistry, Proc. “National Workshop on Atmospheric Chemistry,” NWAC 99 (eds. P. C. S. Devara and P. Ernest Raj), IITM, Pune, India. Singh S. K. and France-Lanord C. (2002) Tracing the distribution of erosion in the Brahmaputra watershed from isotopic compositions of stream sediments. Earth Planet. Sci. Lett. 202, 645– 662. Singh S. K., Trivedi J. R., Pande K., Ramesh R., and Krishnaswami S. (1998) Chemical and Sr, O, C, isotopic compositions of carbonates from the Lesser Himalaya: Implications to the Sr isotope composition of the source waters of the Ganga, Ghaghara and the Indus Rivers, Geochim. Cosmochim. Acta 62, 743–755. Singh S. K., Reisberg L., and France-Lanord C. (2003) Re-Os isotope systematics of sediments of the Brahmaputra River system, Geochim. Cosmochim. Acta 67, 4101– 4111. Solomon D. K., and Cerling T. E. (1987) The annual carbon dioxide cycle in a montane soil: Observations, modeling and implications for weathering. Water Resource Res. 23, 2257–2265. Stallard R. F. (1995) Tectonic, environmental and human aspects of weathering and erosion: A global review using a steady-state perspective. Ann. Rev. Earth Planet. Sci. 23, 11–39. Thakur V. C. (1986) Tectonic zonation and regional framework of eastern Himalaya. Science de la Terre 47, 347–360. Velbel M. A. (1993) Temperature dependence of silicate weathering in nature: How strong of a negative feedback on long term accumulation of atmospheric CO2 and global greenhouse warming? Geology 21, 1059 –1062. White A. F. and Blum A. E. (1995) Effects of climate on chemical weathering in watersheds. Geochim. Cosmochim. Acta 59, 1729 – 1747. APPENDIX Ca Flux of Siang at Pasighat ⫽ 3 ⫻ 106 t y⫺1 Dibang at Sadeya ⫽ 0.3 ⫻ 106 t y⫺1 Lohit at Sadeya ⫽ 0.7 ⫻ 106 t y⫺1 Total Ca flux ⫽ 4 ⫻ 106 t y⫺1 (i) Ca flux leaving Dibrugarh ⫽ 7 ⫻ 106 t y⫺1 (ii) Ca flux gained between Pasighat, Sadeya and Dibrugarh ⫽ ⬃3 ⫻ 106 t y⫺1 (iii) Particulate Ca flux (following data from Singh and France-Lanord, 2002) Detrital carbonate (calcite and dolomite) content of Siang at Pasighat ⫽ 4 wt% Dibang at Sadeya ⫽ 2 wt% Lohit at Sadeya ⫽ 3 wt% Brahmaputra at Dibrugarh ⫽ ⬃0 wt% Flux of sediment from Siang ⫽ ⬃50% Flux of sediment from Dibang and Lohit ⫽ ⬃10% Total sediment flux from the Brahmaputra ⫽ 700 ⫻ 106 t/yr Ca flux at Dibrugarh gained due to dissolution of detrital carbonate during transport from Pasighat and Sadeya to Dibrugarh ⫽ ⬃3– 4 ⫻ 106 t y⫺1 (iv) Therefore, from (i), (ii), (iii), and (iv), Ca flux leaving Dibrugarh ⬇ Total Ca flux coming from Siang, Dibang, and Lohit ⫹ Ca flux gained due to dissolution of detrital carbonate between Pasighat, Sadeya, and Dibrugarh.