Journal of Hydrology 391 (2010) 1–8

Contents lists available at ScienceDirect

Journal of Hydrology

journal homepage: www.elsevier .com/ locate / jhydrol

Stable isotope ratios in precipitation and their relationship with meteorologicalconditions in the Kumaon Himalayas, India

U. Saravana Kumar a,*, Bhishm Kumar b, S.P. Rai b, Suman Sharma a

a Isotope Hydrology Section, Isotope Applications Division, Bhabha Atomic Research Centre, Mumbai 400 085, Indiab Hydrological Investigations Division, National Institute of Hydrology, Roorkee 247 667, India

a r t i c l e i n f o

Article history:Received 29 July 2009Received in revised form 21 April 2010Accepted 10 June 2010

This manuscript was handled by L. Charlet,Editor-in-Chief, with the assistance of PeterWolfgang Swarzenski, Associate Editor.

Keywords:Environmental isotopesSource of moistureAltitude effectKumaon Himalayas

0022-1694/$ - see front matter � 2010 Elsevier B.V. Adoi:10.1016/j.jhydrol.2010.06.019

* Corresponding author. Tel.: +91 022 25592418; faE-mail addresses: uskumar1@rediffmail.com, vs

Kumar).

s u m m a r y

For the first time, environmental isotopic (d2H, d18O, 3H) data, predominantly based on precipitation sam-ples in the Kumaon Himalayas, India, have been used to understand the influence of various meteorolog-ical factors governing rainout processes in the region. Further, the data are also used to understand theorographic effects in precipitation and to estimate the altitude effect in stable isotopic ratios in precipi-tation, etc.

The interpretation of the isotopic data revealed that the source of moisture for winter (October–Febru-ary) and summer (May) precipitation in the Kumaon Himalayas is mainly from the Western Disturbanceswhereas for the remaining period the source is monsoonal (southwest). The d2Y�d18O relationship in thelocal precipitation during the monsoon season shows a distinct seasonal effect, with a slope of 7.6. Thewinter and summer precipitation samples measured higher environmental 3H compared to southwestmonsoon samples, thus indicating continental evaporated moisture. There is wide range of altitudeeffects (d18O variation per 100 m elevation: �0.30‰ [July–August]; �0.57‰ [September]) with the meanaltitude effect being �2.61‰ and�0.36‰ per 100 m elevation for d2H and d18O respectively. These valuesare different from that reported earlier for the region based on the isotopic compositions of springs/riverssamples. The ‘altitude effect’ in successive precipitation is basically a temperature dependent phenome-non and is explained on the basis of adiabatic cooling related rainout process (dry adiabatic lapse rate,moist adiabatic lapse rate and saturated adiabatic lapse rate for moist air mass). The altitude effect isfound to be more reliable in case of d2Y, as deuterium is least affected by secondary evaporation. Theeffect of secondary evaporation has been observed on the true ‘‘altitude effect”. Secondary evaporationof rainfall increases the oxygen isotopic ratios and the increase is directly proportional to the vertical dis-tance travelled by the raindrops through air.

� 2010 Elsevier B.V. All rights reserved.

1. Introduction

The Himalayas, the youngest mountain range on the earth, givesrise to three of the world’s major river-systems viz. the Indus, theGanges and the Brahmaputra. In spite of the hydrological impor-tance of the region, very few studies have been reported on stableisotopic (d2Y and d18O) characteristics in precipitation in the re-gion. Evidently, none of the studies on Himalayas (e.g., Bahadur,1976; Ramesh and Sarin, 1992; Bartarya et al., 1995) has presentedthe actual observed altitude effect in precipitation. Thus the com-plexity in the analysis of the relationship between isotopic compo-sition of precipitation and elevation in parts of Himalayas has beenreported by Poage and Chamberlain (2001).

ll rights reserved.

x: +91 022 25505151.k@barc.gov.in (U. Saravana

For the first time, environmental isotopic data predominantlybased on precipitation samples, and some water bodies in the lakeNaini Catchment, situated in the Kumaon Himalayas, India, havebeen used to understand the various meteorological factors gov-erning rainout processes in the region and the orographic effectsin precipitation and estimate the altitude effect in precipitationusing actual observations (air temperature, relative humidity andrainfall etc.).

2. The study area

The study area (Fig. 1), situated in the Kumaon Himalayas (Lat.29�240N; Long. 79�230E; Area: �4.7 km2), encompasses an altituderange between 1937 and 2600 m msl. The mean annual rainfall inthe study area is 2488 mm. The mean monthly rainfall, relativehumidity and air temperature recorded at lake Naini site for1995 are shown in Table 1. The rainfall amount, in the Himalayanranges, generally decreases westward because of the increasing

70° 80° 90°20°

30°

UP

UK

0 500 m

Fig. 1. The location and drainage map of the study area.

Table 1The monthly rainfall, relative humidity, air temperature, d2H, d18O and deuteriumexcess (d) values of precipitation at the Lake Site in the study area in the year 1995.

Month Rainfall(mm)

Relativehumidity (%)

Airtemperature(�C)

d2H(‰)

d18O(‰)

d (‰)

January 46.0 58 4 �90 �12.0 +6.3February 59.3 59 6 �83 �11.3 +7.2March 62.3 48 10 �77 �10.5 +6.9April 5.8 33 15 �7 �1.6 +6.0May 18.5 41 20 �64 �10.5 +19.9June 89.0 69 22 �64 �10.5 +19.9July 509.0 82 19 �79 �11.4 +12.0August 632.0 86 18 �83 �11.3 +7.2September 341.4 78 17 �77 �10.5 +6.9October 0.0 73 15 – – –November 0.0 51 12 – – –December 10.2 54 9 �88 �12.6 +12.7

2 U. Saravana Kumar et al. / Journal of Hydrology 391 (2010) 1–8

distance from the main source of moisture, i.e., Bay of Bengal. Theprecipitation during the monsoon season (June–September) isreceived as moderate to heavy. In contrast to this, during winterseason (January–March) the precipitation is light to moderate(62.3–46.0 mm) with occasional snowfall, caused by extratropicalweather systems of mid-latitude regions (originating over CaspianSea). These winter weather systems are known as Western-disturbances.

The study area forms a synclinorium which is cut diagonallyinto two parts by the Naini Fault. The northwestern part is madeup exclusively of argillaceous limestone and marlites whereasthe southwestern part comprises of dolomite with limestone andblack carbonaceous slates (Valdia, 1988). As shown in Fig. 1, thereare a number of streams and springs in the study area. Most of thestreams in the region are ephemeral while the perennial ones aremostly fed by springs.

The drainage pattern in the study area is controlled by mechan-ical structures. As shown in the drainage map of study area (Fig. 1),

U. Saravana Kumar et al. / Journal of Hydrology 391 (2010) 1–8 3

the area to the west of Lake Naini is marked by very few streamswhile the area to the east is characterised by several parallel tosub-parallel streams. Rawat (1987) ascribed this parallel drainagein the east due to the presence of several fractures, faults andjoints, and the absence of a dense drainage network in the westdue to the presence of dolomites and limestone of Upper Krol.Out of 20 channels draining the basin, only one is perennial (Nain-adevi Drain), whose flow is sustained mainly by discharge from ahigher altitude spring.

3. Methods

A number of precipitation, springs and streams samples werecollected in 1994–1995 for environmental isotopic analyses (d2H,d18O, 3H). The precipitation samples were collected from four sta-tions set up at different altitudes, among which three were set-up within the basin and one on the downstream side of the NainiLake Basin. The samples were collected using plastic containers

S5

S6

S7

S8

S9S10

S11

1733

2079

2434

2352

2375

2610

Government Hous

Kumaun University

Nainital Politechnic

Pardha Dhara

Dr. House

Nanda Devi

Lake S

Pashan Devi

22002100

2000

2100

2200

1600

1500

1400

1300

1

Khurpatal

Sukhatal

AliChunnadhara

Tiffin Top

Spring Locations Raingauge Sites

Important Land Marks Hill Top

Fig. 2. The water sampling locations

and metal funnels having effective catch-diameters of 210 mm.The daily precipitation collected was transferred into a sealed con-tainer after careful measurement of the amount of rain. Each sam-ple collected during several rainy events in each calendar monthwas thoroughly mixed to yield a representative integrated monthlysample. The integrated samples, measuring about 60 ml were col-lected for d2H and d18O analyses, in a tightly-sealed screw cap bot-tle (plastic or glass) and waterproof tape bound around the cap andstored in a cooler place to ensure no evaporation takes place. Alsosamples measuring about 1000 ml were collected for environmen-tal 3H analysis. The water sampling locations are shown in Figs. 1and 2.

In-situ parameters of the samples (T, EC, etc.) were measuredusing Cole Parmer conductivity meter. All efforts were made to col-lect samples from various locations during different seasons. How-ever during some months, no sampling was possible, as there wasno flow in the streams and springs except in Nandadevi Drain. Ford2H and d18O analyses, an online preparation system (ISOPREP18)

S1

S2

S3

S4

2397

2481

2121

1968

1959

1951

2107

e

ite

Lake

View

Nainatal

Lake

Observatory

Snow View

2300

2400

2200

2000

1900

1800

800

Rais Hotel

Gupha Mahadev

Sipahldhara

1700

1600

1300

1400

1500

14001300

1500

16001700

na Cottage

2100

2000

19001800

Scale

of the springs in the study area.

δ H = 8 * O + 10

δ H = 7.5*δ

O + 4.8

(r=0

.98, n

=15)

LMWL :

GMWL :

Precipitation Streams Sipahidhara spring Gupha Mahadev spring Dr. House spring Lake View spring Alma Cottage spring

-13 -12 -11 -10 -9 -8 -7 -6 -5 -4 -3 -2 -1

-100-90-80-70-60-50-40-30-20-10010

(‰)

δ2H

(‰)δ18O

Fig. 3a. The d2H–d18O plot of the springs, streams and precipitation in the studyarea.

-13 -12 -11 -10 -9 -8 -7 -6 -5 -4 -3 -2 -1

-100

-90

-80

-70

-60

-50

-40

-30

-20

-10

0

10

(‰)

(‰)

H

18O

GMWL (Craig, 1961) LMWL (Lake Naini; Present study) LMWL (New Delhi; Datta et al., 1991) LMWL (Gaula river catchment; Bartarya el al., 1995) LMWL (Ganges river headwaters; Ramesh and Sarin, 1992)

Fig. 3b. The LMLs of deduced from the earlier investigations in the region alongwith that of the nearest IAEA/MWO GNIP station, namely New Delhi, and theGMWL.

4 U. Saravana Kumar et al. / Journal of Hydrology 391 (2010) 1–8

and 602 E Auto Mass Spectrometer supplied by VGISOGAS, UKwere used. The zinc method (Navada, 1988) was used in the pres-ent study for d2H analysis which is simple and avoids the memoryeffect inherent in the uranium method which was used earlier.d18O in the samples was measured using the standard procedure(Navada, 1988). Calibration of d2H and d18O measurements wascarried out using IAEA standards SMOW, GISP and SLAP. The re-sults are reported in the standard d-notation w.r.t SMOW (preci-sion: ±1‰ and ±0.2‰, for d2H and d18O respectively). Forenvironmental 3H analysis, ultra-low level liquid scintillationcounter (Wallac Quantulus) was used. The precision of measure-ment was ±0.5 TU after electrolytic enrichment of the water sam-ples (Navada, 1988).

The meteorological data such as temperature, relative humidityand rainfall were observed at the study area. Precipitation datawere collected by installing a self recording (siphon type) raingauge and an Ordinary Rain Gauge (ORG) within the lake basinand another ORG installed at 2 km south of the lake basin.

4. Results and discussion

4.1. Precipitation

Meteoric water line is the best fit line of the d2H and d18O con-tent of precipitation. Craig (1961) after a global survey of stableisotope contents in precipitation proposed a Global MeteoricWater Line (GMWL), which was later modified by Rozanski et al.(1993) and expressed as:

d2H ¼ 8:17 � d18Oþ 10:35 ðn ¼ 206; r2

¼ 0:98; significant at 0:05 levelÞ ð1Þ

The above GMWL is essentially a global average of several LocalMeteoric Water Lines (LMWL). The equation for the long-term(1961–1978) weighted mean isotopic ratios for monsoon months(June–September) of the nearest IAEA/GNIP station, namely NewDelhi (Datta et al., 1991) is:

d2H ¼ 6:8 � d18Oþ 0:7 ð2Þ

The variation in the slope of the meteoric water line as well asintercept (denoted by ‘d’ or d) clearly indicates the fact that therelationship between d2Y and d18O is more complex in case of In-dian monsoon precipitation. The d2H and d18O values of themonthly precipitation samples collected from the study area areshown in Fig. 3a–b and Tables 1 and 2. Seasonal variation inthe isotope values is clearly seen with enriched values in Aprilcompared to other months (Fig. 3a). The LMWL has the followingrelation

d2H ¼ 7:5ð�0:4Þ � d18Oð�4:0Þ ðr2 ¼ 0:98; n ¼ 15Þ ð3Þ

Eq. (3) can be compared with Eq. (1). The difference in the slopeand intercept of the two equations (Fig. 3b) can be explained, onlyif the effect of secondary evaporation during rainfall is considered,which is quite likely at lower latitudes with higher ambient tem-perature (Krishnamurthy and Bhattacharya, 1991). In comparisonto Eq. (2), the obtained LMWL (Eq. (3)) shows higher slope andintercept (Fig. 3b). This is because the IAEA/GNIP station (New Del-hi) located in the plains has semi-arid climate with generally highambient temperature and hence the secondary evaporation is pre-dominant in precipitation in this region unlike the study area. Theslope of Eq. (3) is comparable with that obtained for the Gaula Riv-er Catchment area of the Kumaon Himalayas (Bartarya et al.,1995):

d2H ¼ 7:0 � d18Oþ 15:0 ðr2 ¼ 0:75; significant at 0:05 levelÞð4Þ

and for the other Himalayan region (Ramesh and Sarin, 1992), viz.for the Ganges River Headwaters (Fig. 3b):

d2H ¼ 7:5 � d18Oþ 8:0 ð5Þ

However, the intercept obtained in the present study is lower. Eq.(4) has been derived from samples collected from springs, as aproxy for precipitation, mostly in summer (June–October) season.Since the intercept of Eq. (4) is high (15), the equation probably rep-resents delayed recharge components from winter precipitation,having moisture source due to Western Disturbances from the Med-iterranean region, in the summer months. The lower intercept inthe present study (Eq. (3)), represents samples from all seasonswith at least three different type of moisture source (i.e., southwestmonsoon, its retreat, the Western Disturbances, etc.) and hence amore representative one.

The re-evaporation of falling rain drops, under dry conditions,has an effect of obscuring the seasonal variations (Ehhalt et al.,1963). However, the secondary evaporation of falling raindropshas similar effects in the present case where the moist climaticconditions prevail. Further, this re-evaporation overshadows thetrue altitude effect as discussed in the following section.

Table 2The d2H and d18O values of precipitation in the study area at different attitudes along with its deuterium excess (d) and rainfall (RF).

Site Altitude (m, msl) July 1995 August 1995 September 1995

d18O (‰) d2H (‰) d (‰) RF (mm) d18O (‰) d2H (‰) d (‰) RF (mm) d18O (‰) d2H (‰) d (‰) RF (mm)

Snow View 2275 �12.3 �87 +11.4 458.0 �12.4 �91 +8.6 613.5 �12.2 �84 +14.0 365.9Alma Cottage 2140 �11.9 �84 +11.4 485.0 �11.8 �88 +6.5 626.9 �11.3 �80 +10.4 356.0Lake Site 1940 �11.4 �79 +12.0 509.0 �11.3 �83 +7.2 632.0 �10.5 �77 +6.9 341.4Gupha Mahadev 1830 �10.7 �77 +9.0 513.6 �11.0 �78 +10.0 639.0 �9.5 �70 +5.7 336.5

U. Saravana Kumar et al. / Journal of Hydrology 391 (2010) 1–8 5

4.2. Orographic effects on precipitation

It is a well-known fact that the interaction of weather systemswith topography in mountainous basins results in non-uniformprecipitation. Literature abounds with reports regarding the varia-tion of rainfall amount and intensity with increase in altitude.Singh et al. (1995) studied the orographic effect of precipitationin different ranges of Western Himalayas observed that there is aremarkable difference in the variation of rainfall with altitudeamong the three ranges viz., Outer, Middle and Greater Himalayas.On the rainfall distribution in the Outer Himalayas, Singh et al.(1995) observed that during monsoon season the rainfall in thewindward side increases up to an altitude of 600 m and then de-creases, while during other seasons the amount of rainfall in-creases linearly with altitude. They attributed this differentialbehaviour to the difference in the weather conditions particularlymoisture content in air, as during monsoon the moisture contentin atmosphere and clouds remains in saturated condition andclouds become mature at relatively lower elevation.

In the present study, monthly rainfall data of four stations lo-cated at different altitudes have been used to analyze the oro-graphic effect in precipitation. The results indicate that therainfall during July and August showed a decreasing trend withaltitude while the rainfall during September showed an increasingtrend with altitude (Table 2). This observation was substantiatedby the variation in the 30-year mean monthly relative humidity(RH) recorded at one of the station. The mean RH is higher duringthe months of July and August and relatively lower duringSeptember.

δ H = -0.0233 (Altitude) - 34.0 (R = 1.00) - Jul.

δ H = -0.0269 (Altitude) - 29.7 (R = 0.97) - Aug.

δ H = -0.0271 (Altitude) - 22.3 (R = 0.92) - Sep. -100

-95

-90

-85

-80

-75

-701700 1900 2100 2300 2500

Altitude (m)

Jul Aug Sep

δ2 H

Fig. 4. The variation in the d2H of precipitation with elevation during differentmonths in the study area.

4.3. Deuterium excess

Deuterium excess (d) in precipitation is a useful tracer of vapoursource and has been related to meteorological conditions at vapoursource regions (e.g., Rozanski, 2005) and to vapour recycling overthe continents (e.g., Gat et al., 1994). The d is computed as;

d ¼ d2H� 8 � d18O ð6Þ

which is the intercept on y-axis (d2H) in the d2H–d18O plot. The dindicates the effect of secondary evaporation. The global mean va-lue of d is approximately +10‰. The calculated d values of the pre-cipitation samples from the study area are given in Tables 1 and 2along with d2H and d18O values of the precipitation. The d parame-ter has origin in contributions of kinetic factors in the overall pre-cipitation process, i.e., evaporation and condensation. Since theisotopic composition of moisture source controls the d value, itshould be independent of altitude as long as secondary evaporationis not significant. In July and August this condition holds good, asthe RH is high but in September the situation is slightly different be-cause RH is lower (Table 2) and possibility of secondary evaporationof precipitation is higher. Therefore, d values are lower in Septem-ber particularly for lower altitudes (1830 m, 1940 m). The higherd values in May (Table 1) could be associated with moisture sourcefrom possible inland sources (Mediterranean, recycled vapour, etc.)

4.4. Altitude effect

It is well-known that precipitation and surface waters becomeincreasingly depleted in d2H and d18O with increasing altitude, calledas ‘altitude-effect’ (Coplen, 1993). The ‘altitude effect’ in successiveprecipitation is basically a temperature dependent phenomenonand is explained on the basis of adiabatic cooling related rainout pro-cess. The adiabatic cooling refers to the change in temperature of amass of air as it moves upward in the atmosphere. Such changes intemperature could be governed by; (a) Dry adiabatic lapse rate forair mass defunct of moisture or (b) Moist adiabatic lapse rate and sat-urated adiabatic lapse rate for moist air mass. The lapse rates varysignificantly with the amount of moisture content of the air mass.This altitude effect has been recognized in almost all the majormountain belts of the world (Poage and Chamberlain, 2001). Thealtitude effect is most often expressed as an isotopic lapse rate andgiven as a per mil (‰) change in d2H and d18O of precipitation per100 m of elevation change. These lapse rates vary significantly withthe amount of moisture content of the air mass. The values are 3–9.78 �C per km ascent of air mass saturated with moisture and dryair (Todd, 2000).

The d2H and d18J values of the monthly precipitation plottedagainst their altitude of sampling sites gave a wide range of alti-tude effect (July and August: �0.30‰, September: �0.57‰, per100 m elevation for d18O; Figs. 4 and 5, Table 3). In Septemberthe RH is lower compared to July and August and hence its moistadiabatic lapse rate is higher. Therefore, the temperature is lowerin September and hence the rainout process occurs at lower tem-perature compared to July and August. This leads to comparativelyhigher fractionation in heavier isotopes. Hence, the altitude effectappears higher in September as compared to that for July and Au-gust. As discussed in the earlier section, the deuterium excess (d)also corroborates the secondary evaporation effect. On the otherhand, the d2H data of September do not show this effect possiblybecause for d2H, the kinetic fractionation factor plays a minor role,while in case of d18O the equilibrium and kinetic fractionation fac-tors have the same magnitude.

δ O = -0.0057 (Altitude) + 0.8 (R = 0.98) - Sep. δ O = -0.0030 (Altitude) - 5.4 (R = 0.98) - Aug.

δ O = -0.0034 (Altitude) - 4.6 (R = 0.96) - Jul.

-15

-14

-13

-12

-11

-10

-9 1700 1900 2100 2300 2500

Altitude (m)

δ18O

Sep Aug Jul

Fig. 5. The variation in the d18O of precipitation with elevation during differentmonths in the study area.

Table 3The estimated altitude effects (per 100 m elevation) in the d2H and d18O values ofprecipitation samples.

Month d18O (‰) d2Y (‰)

July 1995 �0.30 �2.3August 1995 �0.30 �2.7September 1995 �0.57 �2.7

6 U. Saravana Kumar et al. / Journal of Hydrology 391 (2010) 1–8

From the above discussion, it is evident that two different pro-cesses (adiabatic temperature lapse, secondary evaporation) havebearing on the altitude effect in the study area. Therefore, the meanaltitude effect in the rainfall isotope data for the monsoon period(June–September) has been estimated by the rainfall weightedmean method and found to be as given in the following equations:

d2H ¼ �0:0261 � h� 28:7 ðn ¼ 4; r2

¼ 0:98; significant at 0:05 levelÞ ð7Þ

d18O ¼ �0:0036 � h� 4:1 ðn ¼ 4; r2

¼ 0:96; significant at 0:05 levelÞ ð8Þ

Table 4The d2H and d18O values of springs in the study area.

Name of spring (discharge/recharge elevationsa, m msl) December 1994(‰)

d18O d2H

S1 (1800/1833) �10.7 –S2 (1790) �7.5 –S3 (1730) �7.4 –S4 (1720) �7.0 –S5 (1730) �7.7 –S6 (1750/1917) �11.0 –S7 (1640/2127) �11.8 –S9 (1760/1833) �10.7 –S10 (1700/1722) �10.3 –S11 (1650/1890) �10.9 –S12 (1700/1833) �10.7 –Sipahidhara (1880) – �63Gupha Mahadev (1830) �9.5 �64Lake View (2000) – –Pardhadhara (1940) �4.9 �42Dr. House (1960) – –

Name of spring (altitude, m msl) June 1995 August 199

d18O (‰) d2H (‰) d18O (‰)

Alma Cottage (2140) – – –Sipahidhara (1880) �7.3 �48 �6.9Gupha Mahadev (1830) – �50 �7.1Lake View (2000) �8.0 �58 �9.0Pardhadhara (1940) – – �6.1Dr. House (1960) �9.3 �57 �6.4Alma Cottage (2140) – – –

a Recharge elevations estimated for non-perennial springs are based on the altitude e

Therefore, the weighted mean altitude effect in the region isestimated as �2.61‰ and �0.36‰ per 100 m elevation, for d2Hand d18O respectively.

The obtained altitude effect value is different from that reportedfor Gaula river catchment of the Kumaon Himalayas (mean altitudeeffect of �0.14‰ per 100 m elevation for d18O – Bartarya et al.,1995) based on the isotope values of spring samples and compara-ble for the Western Himalayan region (�3.2‰ and �0.38‰ per100 m elevation, respectively for d2H and d18O; Ramesh and Sarin,1992) based on the isotope values of the Ganga river headwaters.Navada (1988) has reported a value of �0.5‰ and �4‰ per100 m elevation, respectively for d18O and d2H in the snow samplesof the Western Himalayas.

In non-Himalayan locations, the literature reported values ofaltitude effects for d18O in precipitation per 100 m elevation are;�0.2‰ in SE California, North America (Friedman et al., 1992),�0.3‰ in Central Chile–South America (Stowhas and Moyano,1993), �0.2‰ in Italian Alps (Dray et al., 1998), �0.2‰ in SouthernAlps, New Zealand (Chamberlain et al., 1999), �0.8‰ in CordilleraBlanca, Peru (Niewodniczanski et al., 1981), �0.9‰ in Central Hin-du Kush, Afghanistan (Niewodniczanski et al., 1981), �1.1‰ in Vic-toria Land, Antarctica (Dansgaard et al., 1973), etc.

4.5. Tritium in precipitation

Tritium in precipitation is by and large of stratospheric origin(Rozanski, 2005). Oceanic and coastal stations show considerablylower concentration than continental stations. The reason for thisis the water vapour exchange between the atmospheric and seasurface, which is perhaps twice as effective in removing tritiumfrom the atmosphere as precipitation alone. In continental area, aconsiderable part of the tritium precipitated from the atmosphereis returned to the atmosphere by evapotranspiration. When com-bined with prevailing wind system this leads to a rather rapid de-crease in tritium concentration over ocean areas downstream of

February 1995(‰)

March 1995 (‰) May 1995 (‰)

d18O d2H d18O d2H d18O d2H

– – Zero discharge– – Zero discharge�7.8 – Zero discharge�7.5 – Zero discharge�7.9 – Zero discharge– – Zero discharge– – Zero discharge– – Zero discharge– – Zero discharge– – Zero discharge– – Zero discharge�6.2 – �7.9 �53 �7.5 –�7.8 �55 �6.8 �44 �8.3 �51– – �5.0 �39 – –– – �7.5 �53 �7.7 �56– – – – – –

5 September 1995 November 1995

d2H (‰) d18O (‰) d2H (‰) d18O (‰) d2H (‰)

– – – – –�45 �8.0 �51 �6.5 �47�51 �8.8 �54 �8.7 �48�60 �5.6 �39 �8.3 �55�44 – – �7.3 �49�43 �10.8 �69 – –– – – �8.2 �53

ffect of d18O in precipitation.

Table 5The environmental tritium content in precipitation in the study area at different altitudes.

Site Altitude (m, msl) October 1994 (TU) February 1995 (TU) July 1995 (TU) August 1995 (TU) September 1995 (TU)

Snow View 2275 – – 6.6 6.6 11.4Alma Cottage 2140 – – 7.1 8.5 11.2Lake Site 1940 17.2 15.3 – 9.5 14.7Gupha Mahadev 1830 – – – 9.5 –

Table 6The d2H and d18O values of streams in the study area.

Name of stream May 1994 October 1994 December 1994 March 1995

d18O (‰) d2H (‰) d18O (‰) d2H (‰) d18O (‰) d2H (‰)

Ramsay Hospital – – – – – �44 Insignificant discharge in the streamsNainidevi �4.7 – �8.5 – �8.7 �60Municipal library – – – – – –

Name of stream May 1995 August 1995 September 1995 November 1995

d18O (‰) d2H (‰) d18O (‰) d2H (‰) d18O (‰) d2H (‰)

Ramsay hospital Insignificant discharge in the streams – �56.3 – �44 �9.6 �64Nainidevi – – �8.2 �48 – –Municipal library – – �7.4 �53 – –Edge Hill – – �8.9 �61 �7.7 –

U. Saravana Kumar et al. / Journal of Hydrology 391 (2010) 1–8 7

coastlines to a more or less stationary level, whereas there will be agradual increase in concentration in continental areas downwindof coastlines.

The primary sources of water for rains over India are the va-pours from the Arabian Sea and the Bay of Bengal or the SouthernIndian Ocean. In the northwestern parts of Himalayas, a distant ori-gin of moisture from Mediterranean Sea is also reported, i.e., calledWestern Disturbances. Recycling through evapotranspiration ofprecipitated water over land can also be a significant source of va-pour. The relative proportion of each varies not only temporallybut also spatially. It is important to identify the source area of pre-cipitated water (i.e., location from where vapours originate) forunderstanding the detailed structure of the regional hydrologicalcycle. The ratio of precipitation that originates over land to the to-tal precipitation amount is termed as ‘continental cycling ratio’.Quantifying the degree of water originated from land is also impor-tant because it gives useful information on possible interactionsbetween land surface hydrology and climate.

The environmental tritium (3H) content in the precipitationsamples (Table 5) at various altitudes range from 6.6 TU (July) to17.6 TU (October). During SW monsoon, the 3H content in precip-itation ranges from 6 to 9 TU whereas during other periods it re-mains between 11 and 17 TU. This probably indicates thedifferent sources of condensing moisture during these months.The high 3H values in September/October and February precipita-tion possibly indicate retreating southwest monsoon and WesternDisturbances with continental evaporated moisture.

4.6. Streams and springs

The isotopic values of the streams fall on the LMWL (Figs. 3aand b, Table 6). Springs and local precipitation feed the streamsas they show intermediate values between spring and local precip-itation, from the isotopic signatures. Seasonality in the isotope val-ues of the higher altitude springs (e.g., Snow View, Alma Cottage,Lake Site, Gupha Mahadev, etc.) is clearly observed. This can beattributed to the fact that since the flow is mainly through frac-tures and fault planes, the springs represent the rainwater samples,with some delay. Also, the slight isotopic enrichment seen in thespring samples compared to the rainwater samples could probably

represent some component of recharge from some of the evaporat-ing surface water bodies seen in the upper reaches of thecatchment.

Based on the computed altitude effects (�2.61‰ and �0.36‰

per 100 m elevation for d2H and d18O, respectively; Fig. 5), the re-charge elevations of some of the non-perennial springs feeding tothe streams of the lake were estimated and given in Table 4. Dueto non-availability of discharge-data and observed seasonal isoto-pic variations of the perennial springs (Table 4), their recharge ele-vations could not be estimated. The environmental 3H contents ofsome of the springs (9.2–12.5 TU in February 1995) suggest thatthey are modern waters and hence the recharge elevations mostprobably lie within the catchment (local recharge) and not outsidethe catchment (remote recharge). This probably corroborates thefindings of Saravana Kumar et al. (2001), who based on a (environ-mental) tritium and water-balance concept developed a model toestimate the residence time of groundwater contributing to awell-mixed and hydrologically steady state lake–groundwater sys-tem. He estimated the residence time of groundwater contributingto lake Nainital to be about 14 years.

5. Conclusions

The oxygen and hydrogen isotope ratios of the precipitation inNaini Lake Basin, located in the Kumaon Himalayas, observed dur-ing 1994–1995 have been analysed. The mean altitude effect in theHimalayan precipitation, based on observed data, has been re-ported for the first time (�2.61‰ for d2Y and �0.36‰ for d16J,per 100 m increase in the altitude). The ‘altitude effect’ in succes-sive precipitation is basically a temperature dependent phenome-non and is explained on the basis of adiabatic cooling relatedrainout processes (dry adiabatic lapse rate, moist adiabatic lapserate and saturated adiabatic lapse rate for moist air mass). A ‘sec-ondary-altitude effect’ a resultant of the secondary evaporation isrecognized, an effect that is distinct from the mean altitude effect.The findings of this study will be useful in surface water–ground-water interaction studies and in identifying recharge areas ofsprings/groundwater sources in the region, especially with sup-porting hydrogeology and hydrogeochemistry.

8 U. Saravana Kumar et al. / Journal of Hydrology 391 (2010) 1–8

Acknowledgements

The authors would like to express their sincere gratitude to Dr.Rm.P. Nachiappan, a former scientist of the National Institute ofHydrology, Roorkee, and Dr. S.V. Navada, formerly Head, IsotopeHydrology Section of Bhabha Atomic Research Centre, Mumbaifor their help and guidance given during the course of study. Sin-cere thanks are due to Dr. V. Venugopal, Director, Radiochemistry& Isotope Group and Dr. Gursharan Singh, Head, Isotope Applica-tions Division, BARC for giving the permission to publish this arti-cle. The authors are highly thankful for the comments andsuggestions of the editor and the anonymous reviewers that helpedus enormously in modifying the article. The authors also expresstheir thanks to Sri A. K. Gupta who helped in improving the styleand presentation of the manuscript.

References

Bahadur, J., 1976. Altitudinal variation in deuterium concentration of water samplesof central Himalayas. Indian J. Meteorol. Hydrol. Geophys. 27, 55–58.

Bartarya, S.K., Bhattacharya, S.K., Ramesh, R., Somayajulu, B.L.K., 1995. d18O and dDsystematics in the surficial waters of the Gaula river catchment area, KumaunHimalaya. Indian. J. Hydrol. 167, 369–379.

Chamberlain, C.P., Poage, M.A., Craw, D., Reynolds, R.C., 1999. Topographicdevelopment of the Southern Alps recorded by the isotopic composition ofauthigenic clay minerals, South Island, New Zealand. Chem. Geol. 155, 279–294.

Coplen, T.B., 1993. Uses of environmental isotopes. In: Alley, W.M. (Ed.), RegionalGroundwater Quality. Van Nostrand Reinhold, New York, pp. 227–254.

Craig, M., 1961. Isotopic variation in meteoric waters. Science 133, 1702–1703.Dansgaard, W., Johnsen, S.J., Clausen, H.B., Gundestrup, N., 1973. Stable isotope

glaciology. Meddelelser om Gronland 197, 1–53.Datta, P.S., Tyagi, S.K., Chandrasekhran, H., 1991. Factors controlling stable isotope

composition of rainfall in New Delhi. J. Hydrol. 128, 223–236.Dray, M., Jusserand, C., Novel, J.P., Zuppi, G.M., 1998. Air mass circulation and the

isotopic ‘‘shadow effect” in precipitation in the French and Italian Alps. In:Proceedings of Symposium on Past and Current Environmental Changes in theHydrosphere and Lithosphere, 1998. IAEA, Vienna, pp. 107–117.

Ehhalt, D., Kott, K., Nagel, J.F., Vogel, J.C., 1963. Deuterium and oxygen-18 in rainwater. J. Geophys. Res. 68, 3775–3780.

Friedman, I., Smith, G.I., Gleason, J.D., Warden, A., Harris, J.M., 1992. Stable isotopecomposition of waters in southeastern California 1. Modern precipitation. J.Geophys. Res. 97, 5795–5812.

Gat, J.R., Bowser, C.J., Kendall, C., 1994. The contribution of evaporation from theGreat Lakes to the continental atmosphere: estimate based on stable isotopedata. Geophys. Res. Lett. 21, 557–560.

Krishnamurthy, R.V., Bhattacharya, S.K., 1991. Stable oxygen and hydrogen isotoperatios in shallow groundwaters from India and a study of the role ofevapotranspiration in the India monsoon. In: Taylor Jr., H.P., O’Neil, J.R.,Kaplan, I.R. (Eds.), Stable Isotope Geochemistry: A Tribute to Samuel Epstein, pp.187–194.

Navada, S.V., 1988. Application of Environmental Isotope Geochemistry inHydrological Studies. PhD. Thesis, University of Bombay.

Niewodniczanski, J., Grabczak, J., Baranski, L., Rzepka, J., 1981. The altitude effect onthe isotopic composition of snow in high mountains. J. Glaciol. 27, 99–111.

Poage, M.A., Chamberlain, C.P., 2001. Empirical relationships between elevation andthe stable isotope composition of precipitation and surface waters:considerations for studies of paleoelevation change. Am. J. Sci. 301, 1–15.

Ramesh, R., Sarin, M.M., 1992. Stable isotope study of the Ganga (Ganges) riversystem. J. Hydrol. 139, 49–58.

Rawat, J.S., 1987. Morphology and morphometry of the Naini lake, Kumaon lesserHimalaya. J. Geol. Soc. India 30, 493–498.

Rozanski, K.L., Araguas-Araguas, Gonfiantini, R., 1993. Isotopic patterns in modernglobal precipitation. In: Swart, P.K. (Eds.), Climate Change in ContinentalIsotopic Records. Geophysics Monograph Series, vol. 78, pp. 1–36.

Rozanski, K., 2005. Isotopes in atmospheric moisture. In: Aggarwal, P.K., Gat, J.,Froehlich, K. (Eds.). Isotopes in the Water Cycle. pp. 291–302.

Saravana Kumar, U., Navada, S.V., Rao, S.M., Bhishm Kumar, Murthy, J.S.R., 2001.Environmental isotope study on hydrodynamics of lake Naini, Uttar Pradesh,India. Hydrol. Process. 15, 425–439.

Singh, P., Ramashastri, K.S., Kumar, N., 1995. Topographical influence onprecipitation distribution in different ranges of Western Himalayas. Nord.Hydrol. 26, 259–284.

Stowhas, L., Moyano, J.C., 1993. Simulation of the isotopic content of precipitation.Atmos. Environ. 27A, 327–333.

Todd, S., 2000. Glossary of Meteorology. American Meteorological Society, Boston.<(http://amsglossary.allenpress.com/glossary)>.

Valdia, K.S., 1988. Geology and Natural Environment of Nainital Hills, KimaonHimalaya. Gyanodaya Prakashan-Nainital, pp. 392.