Advertisement

Natural Hazards

, Volume 92, Issue 2, pp 635–645 | Cite as

Tree ring width variations over western Himalaya in India and its linkage with heat and aridity indices

  • Somaru Ram
Original Paper

Abstract

Tree ring chronologies from different sites of western Himalaya have been carried out in relation to rainfall, temperature, palmer drought severity index, and heat and aridity indices of the region. The first principal component which was developed using the multi-sites chronologies of Himalaya has explained 50% common variance is positively correlated with rainfall, aridity and palmer drought severity index and negatively with temperature and heat index during spring season (February–May). The existence of strong correlation indicates that heat and aridity indices over the region might be one of the important climatic parameters which play the significant role in tree growth process. Particularly, heat index’s influence over the region indicated larger impact on annual ring width patterns than temperature.

Keywords

Heat index Aridity index Western Himalaya Tree rings 

1 Introduction

As the changes in mountain ranges elevation, geographical setting, western disturbance originating over the mediterranean sea, biodiversity, forest type, land use and land cover changes, snow covered, and river network over western Himalaya play the important role in diverse of climate over the region. Also, it makes complication to society in understating of climate variability and change over the region. Even at short distance, climate shows big differences. It shows high variability in vicinity, especially for rainfall.

Besides, El Nino–Southern Oscillation (ENSO) events are one of the important driving factors which influence year-to-year variability in climate over lower and middle latitude of the globe (Whetton and Rutherford 1994). Excess or failure of the south west monsoon rainfall over India can affect millions of people every year as evidenced by Gadgil (1996). The intensity of southwest monsoon rainfall over India is monitored by the thermal contrast between land and ocean (Webstar et al. 1998; Pant and Parthasarathy 1981; Parthasarathy et al. 1987). These episodes involve large-scale ocean–atmosphere interactions.

However, to know the long-term climate variability/change and its possible cause over the region, long-term high-resolution proxy climate records of tree rings have been used by several researchers in relation to climate variability/change for last few centuries (Hughes 1992; Pant et al. 1998, 2000; Borgaonkar et al. 1999; Yadav and Singh 2002; Ram 2012; Ram and Borgaonkar 2014; Singh and Yadav 2005; Singh et al. 2006, 2009; Yadav et al. 1999, 2009, 2004). They showed the potential of trees growth for dendroclimatic analysis and indicated the usefulness of trees growth to supplement the instrumental weather records to past few centuries and millennia in absence of long-term climatic records over western Himalaya (Singh et al. 2006; Yadav et al. 2009, 2004; Hughes 1992; Shah and Mehrotra 2017). Moreover, some other studies based on tree growth–climate relationship from rugged Himalayan region also showed that tree growth is influenced by the moisture availability at root zone of the trees (Ram and Borgaonkar, 2014, 2016, 2017; Cook et al. 2010; Yadav et al. 2015). Ram and Borgaonkar (2014) showed that Himalayan tree growth is more influenced by potential evapotranspiration and vapor pressure of the region than temperature. Overall, they have shown that tree growth over western Himalaya is mainly influenced by rainfall and temperature only so far. In the present study, the attempt has been made to understand the association of tree growth with aridity and heat index, which may provide the valuable information to society for better understanding the long-term climate variability/change during the past several centuries over western Himalaya.

2 Materials and method

2.1 Tree ring data

Multi-sites (Ghansali, Kanasar, Tuni, Dhanolti, and Jageswar) residual tree ring chronologies from western Himalaya have been used from the website, http://www.ncdc.noaa.gov/paleo/treering.html/ to re-investigate tree growth–climate relationship (Fig. 1), because the residual tree ring width chronologies after standardization indicate mainly the high-frequency signal over a large area of the sites. It is found more reliable in the analysis of dendroclimatic studies (Fig. 2) and is supposed to be mostly the effect of climate on tree growth (Shah and Mehrotra 2017; Borgaonkar et al. 1996, 1999, Yadav et al. 1999, 2004, 2009, Ram and Borgaonkar 2014, 2016, 2017; Cook et al. 2010). Tree ring width index chronologies from different sites of western Himalaya are shown in Fig. 2.
Fig. 1

Map showing the study areas of western Himalaya; Triangle: grid point climatic data; Circle: Tree ring width data

Fig. 2

Different site residual tree ring width index chronologies over western Himalaya

Numbers of trees cores samples and summary of tree ring width chronologies statistics for each site are shown in Table 1. Their standardization processes and chronologies summary have been already discussed in Borgaonkar et al. (1999). Cross-correlation among the chronologies for the common period 1845–1988 ranges between 0.20 and 0.50. All correlation coefficients (CCs) are found significant in Table 2. Based on their strong CCs, the principal component analysis has been performed to avoid multi-collinearity in chronologies as evidenced by Ram and Borgaonkar (2014, 2016). The first principal component (PC1) which has eigenvalues 2.4693 explained the highest 50% common variance indicates regionally common forcing (Fig. 3), i.e., climate has been used in the present analysis.
Table 1

Tree ring chronologies statistics

Site name

Series

Length (year)

Species’s name

Location

Total time span

EPS > 0.85a

AR1b

SNR

Ghansali

25

195

Pinus roxburghii

30° 37’N; 78° 45’E

1796–1990

1840

− 0.01

6.0

Kanasar

27

278

Cedrus deodara

30° 45’N; 77° 48’E

1711–1988

1850

0.02

5.7

Tuni

29

188

Pinus roxburghii

30° 50’N; 77° 26’E

1801–1988

1870

0.01

5.6

Dhanolti

12

271

Picea smithiana

30° 45’N; 78° 25’E

1720–1990

1891

0.04

2.5

Jageswar

13

334

Cedrus deodara

29° 46’N; 79° 10’E

1657–1990

1795

0.0

4.6

aThe first year that EPS (expressed population signal) shows greater than 0.85 (Wigley et al. 1984)

bAutocorrelation order 1; SNR : signal-to-noise ratio

Table 2

Correlation coefficients among the site chronologies during 1845–1988

 

Ghansali

Kanasar

Tuni

Dhanolti

Jageswar

Ghansali

1

0.31**

0.20*

0.36***

0.38***

Kanasar

 

1

0.37***

0.48***

0.50***

Tuni

  

1

0.25**

0.24**

Dhanolti

   

1

0.49***

Jageswar

    

1

*Significant at 5% level, **significant at 1% level, *** significant at 0.1% level

Fig. 3

First principal component time series (PC1) over western Himalaya during 1845–1988

2.2 Climate data

The gridded monthly mean temperature and rainfall from climate research unit (CRU) (Mitchell and Jones 2005) have been used in the present analysis to analyze the tree growth–climate relationship. The grid boxes (30.25°N, 77.25°E; 30.75°N, 77.25°E; 30.25°N, 77.75°E; 30.75°N, 77.75°E; 30.75°N, 78.25°E; and 30.25°N, 78.75°E) around the sampling areas have been chosen for tree ring analysis (Fig. 1), as representative of meteorological station in the absence of observed climatic data nearby. Monthly heat and aridity indices were computed by using the empirical formula developed by Thornthwaite (1948) to see the impact of indices on tree growth. In addition to this, monthly self-calibrating palmer drought severity index (scPDSI) of the same grid from CRU has been used in relation to tree growth (Ian et al. 2014). A regional series has been prepared for aridity index, palmer drought severity index, rainfall, temperature, and heat index for the period 1901–2002 by taking mean of all grid boxes (Shah and Mehrotra 2017; Yadav et al. 1999, 2004, 2009; Ram et al. 2008; Ram and Borgaonkar 2014, 2016, 2017). Mean monthly variation of heat and aridity indices over western Himalaya is shown in Fig. 4 during 1901–2002. The highest temperature and heat index in month of June and lowest in month of January have been recorded at study area (Fig. 4).
Fig. 4

Mean monthly variation of rainfall, temperature, heat and aridity index over western Himalaya during 1901–2002

3 Tree growth–climate relationships

Correlation analysis between PC1 and aridity index (AI), palmer drought severity index (PDSI), rainfall (RF), temperature (TM) and heat index (HI) for the common period 1901–1988 has been carried out on monthly basis (Table 3). For analyzing the relationship, monthly AI, PDSI, RF, TM, and HI from previous year’s October (ending of previous year’s growing season) to current year’s October (ending of growing current’s growing season) were used with PC1 (Table 3). AI and PPT showed significant positive relationship with tree growth during current year’s February, March and April, whereas TM and HI showed significant negative relationships during prior year’s November, current year’s February, March, April, May and June (Table 3). PDSI showed significant positive relationship from previous year’s October to current’s year July. Based on the results shown by AI, PPT, TM, and HI to PC1 (Table 3), a season was made from February to May to extract the influence of climate on trees growth variations (Fig. 5).
Table 3

Correlation coefficients between PC1 and regional aridity (AI), precipitation (PPT), palmer drought severity index (PDSI), mean temperature (TM), and heat index (HI) on monthly basis during 1901–1988

 

1

2

3

4

5

6

7

8

9

10

11

12

13

AI

0.0

0.15

0.11

0.10

0.23*

0.20*

0.21*

0.17

0.05

0.0

0.1

0.9

0.8

PDSI

0.21*

0.27**

0.29**

0.28**

0.36***

0.38***

0.49***

0.51***

0.39***

0.29**

0.19

0.14

0.13

PPT

0.0

0.13

0.16

0.11

0.27**

0.22*

0.28**

0.19

0.0

0.01

0.0

0.06

0.05

TM

− 0.10

− 0.28**

− 0.17

− 0.02

− 0.20*

− 0.40***

− 0.42***

− 0.38***

− 0.22*

− 0.7

0.01

− 0.02

0.02

HI

− 0.11

− 0.29**

− 0.15

− 0.02

− 0.21*

− 0.42***

− 0.44***

− 0.39***

− 0.22*

− 0.06

0.01

− 0.02

0.03

1: previous October, 2: previous November, 3: previous December, 4: January, 5: February, 6: March, 7: April, 8: may, 9: June, 10: July, 11: August, 12: September, 13: October

*Significant at 5% level, **significant at 1% level, *** significant at 0.1% level

Fig. 5

Variation of PC1(dash) with a Aridity index (solid line), b palmer drought severity index (solid line), c rainfall (solid line), d temperature (solid line) and e heat index (solid line). CC is the correlation coefficients during 1901–1988

4 Results and discussion

The chronology statistics shown in Table 1 indicate that the EPS, which measures the quality of a chronology, is found to be above 0.85 from AD 1840 onward at Ghansali, from AD 1850 onward at Kanasar, from AD 1870 onward at Tuni, from AD1891 at Dhanolti, and from AD1795 at Jageswar (Table 1). The EPS and signal-to-noise ratio ranges from 2.5 to 6.0 indicate the suitability in climate studies (Wigley et al. 1984). The chronology statistics suggest that there is a moderately strong common signal among the site chronology (Table 1). Based on their strong relationship during 1845–1988 (Table 2), the first principal component (PC1) developed by using the multi-species tree ring chronologies was used in further analysis (Fig. 3).

The correlation analysis carried out between PC1 and climate suggests that TM and HI during February to June over western Himalaya have accelerated transpiration and evaporation, which can cause insufficient moisture to root zone of the trees growth as evidenced by Ram and Borgaonkar (2016) (Table 3), which might reduce the pace during cell formation of trees growth. The study showed that dry spring from March to May is not in favor of the trees growth. Similar observations have been noted earlier by Singh and Yadav (2005) over the western Himalayan region. The increasing TM and HI during previous year’s November and current year’s February can reduce the preserved moisture over the region, which is not found conducive during subsequent growing season of the trees growth as evidenced by Ram and Borgaonkar (2014). During June, the high TM and HI may intensify potential evapotranspiration and cause moisture stress condition at root zone of the trees growth.

In case of AI and PPT, the AI and PPT of current’s year February, March and April has positive relationship with trees growth at sites of moisture stressed. It showed that cool and wet early spring (March to April) over the region is found conducive to trees growth as supported by Singh and Yadav (2005). However, February month’s cool may activate rapidly the cambium and physiological processes during onset of growing season of the trees (Table 3). In case of PDSI, the strong positive relationship between PC1 and PDSI during pre-monsoon months (March to May) and early summer (June to July) indicates the importance of soil moisture availability during growing season and showed better correlation coefficients (CCs) with tree growth than rainfall and AI (Table 3). The PDSI during previous year’s October and winter month plays a significant role in cambium activity and physiological processes of the trees growth. The relationship showed that winter cool and wet conditions favor the tree growth of western Himalaya, whereas dry winter months caused insufficient moisture due to increasing HI/TM over the region which is not found conducive to tree growth before the onset of growing season as evidenced by Ram and Borgaonkar (2017) (Table 3). Based on the positive response given by monthly’s AI and RF to growth of the trees, a season has been formed from February to May to see the impact of seasonal climate on trees growth. The CCs between PC1 and AI, PDSI, RF, TM and HI are 0.40, 0.45, 0.51, − 0.51 and − 0.57. All CCs are significant at 0.1% level during 1901–1988 (Fig. 5).

Moreover, to see the temporal stability for longer time scale between PC1 and AI, PDSI, RF, TM and HI, the 31-year sliding correlation coefficients have been carried out for the period 1901–1988 among them (Fig. 6). In case of HI, the sliding CCs fall between − 0.51 and − 0.70; from 1916 onwards, all CCs are significant at 0.1% level in both cases of Tm and HI, but HI showed better performance with tree growth than temperature (Fig. 6). There is a long-term relationship over space and time between tree growth and TM, and HI. The existence of strong negative CCs between PC1 and HI indicates that the increased HI over the region may accelerate the transpiration and evaporation which results moisture stress situation to surrounding trees growth.
Fig. 6

31-yr sliding correlation coefficients between PC1 and AI (square), PDSI (triangle), RF (cross), TM (circle) and HI (Dash). Solid line parallel to x axis indicates significant at 5% level

In case of AI and PPT, the CCs are observed significantly from 1916 to 1949. Thereafter, the CCs are declining in both the cases (Fig. 6), whereas PDSI showed significant correlation coefficients from 1916 onwards (Fig. 6). However, the reason for declining the CCs is not well known, but the increasing temperatures, potential evapotranspiration and vapor pressure over western Himalaya since around 1936 might be the reason which may enhance transpiration and evaporation over the region which cause insufficient moisture to decline the relationship between tree growth and climate (Ram and Borgaonkar 2014, 2013, 2016; Borgaobkar et al. 2011). In addition to this, large-scale land transformation and disturbance in natural resources during recent few decades might be the responsible factors to alter the tree growth–climate relationship as evidenced by Tian et al. (2014).

5 Conclusions

The relationship between PC1 and climatic parameters showed that tree growth over the western Himalaya is influenced by not only rainfall and temperature but also other climatic parameters which might be responsible to carry out the variations in tree growth patterns. Based on the response given by climatic parameters to tree growth, it is observed that HI, PDSI and AI over the western Himalaya are the important climatic parameters which greatly affect the tree growth. More than 33% variance in tree rings explained by HI only shows that multi-sites tree ring chronologies would help in developing of robust climatic reconstruction over western Himalaya.

However, to understand the long-term effect of HI, PDSI and AI on tree growth, the longer tree ring chronologies with good replication of tree core samples from western Himalaya may help in understanding of long-term contemporary climatic variability/change. Such high-resolution proxy climate records other than temperature and rainfall may be useful to society for improving their understanding on climate variability/change during the past several centuries.

Notes

Acknowledgements

The author is gratitude to Prof. Ravi S. Nanjundiah, Director, IITM, Pune, and Dr. R. Krishnan, Executive Director, CCCR, IITM, Pune, for their kind support and providing infrastructure facilities. The author is thankful to NOAA for providing tree ring data. The author is thankful to CRU for making climatic data available on the website. The author is also thankful to Dr S.S. Nandargi, Scientist-D, IITM, Pune, for preparing Fig. 1 of the study area.

References

  1. Borgaobkar HP, Sikder AB, Ram S (2011) High altitude forest sensitivity to recent warming: a tree-ring analysis of conifers from western Himalaya, India. Quat Int 236:158–166CrossRefGoogle Scholar
  2. Borgaonkar HP, Pant GB, Rupakumar K (1996) Ring-width variations in Cedrus deodara and its climatic response over the western Himalaya. Int J Climatol 16:1409–1422CrossRefGoogle Scholar
  3. Borgaonkar HP, Pant GB, Rupa Kumar K (1999) Tree-ring chronologies from western Himalaya and their dendroclimatic potential. IAWA 20(3):295–309CrossRefGoogle Scholar
  4. Cook ER, Anchukaitis KJ, Buckley BM, D’Arrigo RD, Jacoby GC, Wright WE (2010) Asian Monsoon failure and megadrought during the last millennium. Science 328:486–489CrossRefGoogle Scholar
  5. Gadgil S (1996) Climate change and agriculture—an Indian perspective. In: Abrol YP, Gadgil S, Pant GB (eds) Climate variability and agriculture. Narosa, New Delhi, India, pp 1–18Google Scholar
  6. Hughes MK (1992) Dendroclimatic evidence from the western Himalaya. In: Bradley RS, Jones PD (eds) Climate since AD1500. Routledge, LondonGoogle Scholar
  7. Ian H, Jones PD, Osborn TJ, Lister DH (2014) Updated high resolution grid of monthly climatic observations—the CRU TS3.10 dataset. Int J Climatol 34:623–642CrossRefGoogle Scholar
  8. Mitchell TD, Jones PD (2005) An improved method of constructing a database of monthly climate observations and associated high resolution grids. Int J Climatol 25:693–712CrossRefGoogle Scholar
  9. Pant GB, Borgaonkar HP, Rupakumar K (1998) Climatic signals from tree-rings: a dendroclimatic investigation of Himalayan spruce (Picea smithiana). Himalayan Geology 19(2):65–73Google Scholar
  10. Pant GB, Parthasarathy B (1981) Some aspect of an association between the southern Oscillation and the indian summer. Arch Meteorol Geophys Bioclimatol B 29:245–252CrossRefGoogle Scholar
  11. Pant GB, Rupa Kumar K, Borgaonkar HP, Okada N, Fujiwara T, Yamashita K (2000) Climatic response of cedrus deodara tree-ring parameters from two sites in western Himalaya. Can J For Res 30:1127–1135CrossRefGoogle Scholar
  12. Parthasarathy B, Sontake NA, Munot AA, Kothawale DR (1987) Drought/floods in summer monsoon sesaon over different meteorological subdivisions of Indiafor the period 1871–1984. J Climatol 7:57–70CrossRefGoogle Scholar
  13. Ram S (2012) Tree growth-climate relationships of conifer trees and reconstruction of summer palmer drought severity index at Pahalgam in Srinagar, India. Quat Int 254:152–158CrossRefGoogle Scholar
  14. Ram S, Borgaonkar HP (2013) Growth response of conifer trees from high altitude region of Western Himalaya. Curr Sci 105(2):225–231Google Scholar
  15. Ram S, Borgaonkar HP (2014) Tree-ring analysis over the western Himalaya and its long-term association with vapor pressure and potential evapotranspiration. Dendrochronologia 32:32–38CrossRefGoogle Scholar
  16. Ram S, Borgaonkar HP (2016) Reconstruction of heat index based on tree-ring width records of western Himalaya in India. Dendrochronologia 40:64–71CrossRefGoogle Scholar
  17. Ram S, Borgaonkar HP (2017) Moisture index during the last two centuries inferred from tree growth in the western Himalaya. Curr Sci 112(12):2453–2455CrossRefGoogle Scholar
  18. Ram S, Borgaonkar HP, Sikder AB (2008) Tree-ring analysis of teak (Tectona grandis L.F.) in central India and its relationship with rainfall and moisture index. J Earth Syst Sci 117(5):637–645CrossRefGoogle Scholar
  19. Shah SK, Mehrotra N (2017) Tree-ring studies of Toona ciliate from subtropical wet hill forest of Kalimpong, eastern Himalaya. Dendrochronologia 46:46–55CrossRefGoogle Scholar
  20. Singh J, Yadav RR (2005) Spring precipitation variations over the western Himalaya, India, since A.D. 1731 as deduced from tree-rings. J Geoph Res.  https://doi.org/10.1029/2004jd004855 Google Scholar
  21. Singh J, Park WK, Yadav RR (2006) Tree-ring based hydrological records for western Himalaya, India, Since AD 1560. Clim Dyn 26:295–303CrossRefGoogle Scholar
  22. Singh J, Yadav RR, Wilmking M (2009) A 694-year tree-ring based rainfall reconstruction from Himachal Pradesh. Clim Dyn, India.  https://doi.org/10.1007/s00382-09-0528-5 Google Scholar
  23. Thornthwaite CW (1948) An approach towards a rational classification of climate. Geogr Rev 38:55–94CrossRefGoogle Scholar
  24. Tian H, Banger K, Bo T, Dadhwal VK (2014) History of land use in India during 1880–2010: large-scale land transformations reconstructed from satellite data and historical archives. Global Planet Change 121:78–88CrossRefGoogle Scholar
  25. Webstar PJ, MaganaV O, Palmer TN, Shukla J, Tomas RA, Yanai M, Yasunari T (1998) Monsoon: processes, Predictability, and the prospectus for prediction. J Geophys Res 103:14451–14510CrossRefGoogle Scholar
  26. Whetton P, Rutherford J (1994) Historical ENSO teleconnections in the eastern hemisphere. Clim Change 2:221–253CrossRefGoogle Scholar
  27. Wigley TML, Briffa KR, Jones PD (1984) On the average value of correlated time series, with applications in dendroclimatology and hydrometeorology. J Clim Appl Meteorol 23:201–213CrossRefGoogle Scholar
  28. Yadav RR, Singh J (2002) Tree-ring based spring temperature patterns over the past four centuries in western Himalaya. Quat Res 57:299–305CrossRefGoogle Scholar
  29. Yadav RR, Park WK, Bhattacharyya A (1999) Spring temperature variations in western Himalaya, India, as reconstructed from tree-rings: AD 1390–1987. Holocene 9:85–90CrossRefGoogle Scholar
  30. Yadav RR, Park WK, Singh J, Dubey B (2004) Do the western Himalaya defy global warming. Geophys Res Lett 31:L17201.  https://doi.org/10.1029/2004GL020201 CrossRefGoogle Scholar
  31. Yadav RR, Braeuning A, Singh J (2009) Tree-ring inferred summer temperature variations over the last millennium in western Himalaya. Clim Dyn, India.  https://doi.org/10.1007/s00382-009-0719-0 Google Scholar
  32. Yadav RR, Misra KG, Yadav AK, Kotlia BS, Misra S (2015) Tree-ring footprint of drought variability in last 300 years over Kumaun Himalaya, India and its association with crop productivity. Quat Sci Rev 117:113–123CrossRefGoogle Scholar

Copyright information

© Springer Science+Business Media B.V., part of Springer Nature 2018

Authors and Affiliations

  1. 1.Indian Institute of Tropical MeteorologyPuneIndia

Personalised recommendations