SNOW AND GLACIER MELT SIMULATION FOR HYDROLOGY IN A TYPICAL HIMALAYAN WATERSHED
by
ANAND VERDHEN
Department of Civil Engineering
Submitted
In fulfillment of the requirements of the degree of
Doctor of Philosophy
to the
INDIAN INSTITUTE OF TECHNOLOGY DELHI JULY 2013
ACKNOWLEDGEMENTS
With all regards, I express my sincere gratitude and appreciation to my supervisors, Prof. B. R. Chahar, Department of Civil Engineering and Prof. O. P. Sharma, Centre of Atmospheric Sciences, Indian Institute of Technology Delhi for their expert guidance, constant encouragement, unconditional trust, support and constructive criticism during the course of my research work. I am deeply indebted to them for their suggestions during the preparation of the manuscript of this thesis and help to achieve my goal.
I express my special regards and thanks to my Student Research Committee members Prof. Mukesh Khare, Prof A. K. Keshari and Prof. Krishna A. Rao for their critical review, valuable suggestions and support to this research work.
I express my sincere thanks to the Authorities of IIT Delhi in general and the Department of Civil Engineering and Centre for Atmospheric Sciences in particular for providing facilities to carry out the research work.
I gratefully acknowledge the help of Prof. A. K. Gosain and Dr. R. Khosa for initiative to me in the Ph.D programme; and the faculty members Prof. N. K. Garg, Prof. A. K. Nagpal, Prof. A. K. Jain, Prof Shashi Mathur, Prof. H. C. Upadhayaya, Prof. Manju Mohan, Prof. M. R. Ravi, Prof. S. Satya, Dr. D. R. Kausal, Dr. S. K. Deb, Dr. Arun Kumar and Dr. Dhanya C. T. for their support and encouragement.
I am thankful to the Authorities of DRDO, Snow and Avalanche Study Establishment (Manali), Chandigarh; Bhakhra Beas Management Board (Pandoh), Chandigarh and National Institute of Hydrology, Roorkee for providing the study data. I am thankful to Mr. A. Ganju, Director, SASE, Dr. M. R. Bhutiyani, Mr. N. K.
Thakur, Mr. Snehmani, Mr. Trilok Das of SASE; Mr. Ashok Gupta and Mr. Handa of
iii
BBMB and Centre for Atmospheric Sciences, IIT Delhi for providing the required data for the study. I wish to thank Dr. K. K. S. Bhatia, Mr. R. D. Singh, Dr. Pratap Singh, Dr. Manohar Arora and other colleagues of NIH Roorkee for extending their help and support.
My special thanks goes to my well wishers and mentor Shri N. Mohan Rao, former Director, SASE, Prof. G. C. Mishra, Dr. Amrendra Prasad, Dr. Satrughan Prasad, IAS Arun Kumar, Mr. Binoy Kumar, Mr. G. N. Thakur, Mr. B. K.
Bandopdhyay, Mr. Ashok Singh, Mr. Vinod Kumar, Mr. Sanjit, Mr. Solanki, Mr.
Vijoy, Dr. Madhusudan and many more, Authorities of Centre for Water Resources Studies, NIT Patna, P.U., ICT Pvt. Ltd., CES(I) Pvt. Ltd. and Redecon (I) Pvt. Ltd.
This acknowledgement will remain incomplete unless I thank research scholars: Shailendra Jain, Shishir Gaur, Tarkeshwar, Aniket Chakrovarty, Raktim Haldhar, Pratikcha, Sudhir, Maheshwaran, Sarvesh, Sridebi Basu, SK Sohani, Boni Narsimlu, Pawan Kumar, Deepak Kumar, Sushmita, Jaishri Patel, Joyti, and staff members: Dr. M. M. Rao, Mr. Chakresh, Mr. Vikram Chand, Mr. Rajbir Agrawal, Mr. N. R. Gahlot, Mr. Amit, Mr. Mani Ram for utmost cooperation at every stage, good ideas and humour to keep the study period pleasant.
I am grateful to my parent, family and children who accepted my challenging task at the this stage and lastly Shri Brahmdeo Singh, Alok, Aarsh Verdhen, Himritu, Sanjay, my late mother, grandmothers and almighty GOD shall remain the ideal to take all the credit.
New Delhi, July 2013 Anand Verdhen
iv
ABSTRACT
The Himalayan watershed occupies a prime position among the water resources with naturally regulated contribution from snow and ice melt that makes its streams perennial. In spite of the great significance of the region, it lacks adequate observational network. Daily, weekly or monthly snowmelt flow in non-monsoon periods are critical for the livelihood and life of about a billion people living in the basin of the Himalayan rivers. The masses mainly depend on these rivers for water supply, irrigation, hydro-power and other water needs. The hydrology of the Himalyan rivers has been investigated extensively yet the snowmelt simulation deficiency still remains as a major issue as the existing literature indicates. Moreover, knowledge about spatiotemporal characteristics and variability of temperature, precipitation, snowpack, snow cover and snowmelt are essential to model the hydrological processes. The snow and glacier melt component analysis and simulation under varying climatic conditions could improve the predictability of snowmelt and runoff. This study is based on the observations from three snow-meteorological stations, viz., Bhang, Solang and Dhundi and discharge data from Palchan/Manali Bridge gauge station for the Beas watershed, and short term research expedition to Chhota Shigri glacier on Pir-Panjal range of the western Himalayas in Himachal Pradesh. The snowpack melt dominates the flow, after the second week of February, at the study data stations and it happens so in July at glaciers. Therefore, the springtime weekly, daily and seasonal data were analyzed to develop the physical relations and models for the ablating snow cover.
v
The Temperature Lapse Rate (TLR) is an important parameter that is a critical part for the spatial projection of temperature, which has been determined from linear
vi
regression relation of seasonal and springtime weekly maximum, minimum, mean and mixed temperatures. Surprisingly, the TLR for the lower section BS (Bhang to Solang) lies between 0.6 and 1.9 °C/100 m. That means it goes up to nearly twice the adiabatic TLR. But the TLR in section BD (Bhang to Dhundi) is found to lie between 0.34 and
1.28 °C/100 m whereas in section SD (Solang to Dhundi) it remains normal (i.e.
between 0.10 and 0.92 °C/100 m). The stratified TLR methodology has been used to identify a value 0.575 °C/100 m through simulation for the temperature projection on the base mean temperature up to 12 °C, while a value of 0.64 °C/100 m has been determined for the base mean temperature above 12 °C. The average error, range of the error and standard error of the estimates are -0.06, ±3.5 and 1.8 °C, respectively.
The seasonal data indicate rise in Tmax by 0.1 oC/yr. On the whole, the inter-station and inter-annual cross-relation signifies a rate of rise in Tmin of the order of 0.0027 oC/100 m/yr only. However, the measurement accuracy in temperature is up to two decimal places.
Other important variables viz., weekly rainfall, precipitation (snowfall and rainfall), snowfall and snow depth have been analyzed with respect to air temperatures of Bhang to obtain spatial relations and inter-annual variability. The degree-day melt in water equivalent varies between 2 and 11.5 mm °C-1 d-1 though it may rise up to 13 mm °C-1 d-1 for non-zero snow condition. Temperature, snowfall, rainfall and snow depth correction per 100 m of rise in elevation have been estimated as -1.09 °C, 31.2 cm, -7.72 mm and 27.95 cm, respectively. The critical temperatures for snowfall, rainfall and equilibrium conditions have been determined. The snowfall and rainfall mixed precipitation occurs within 0.65 and 11.5 °C of weekly mean temperature for which distribution pattern has been developed. This snowmelt factor is temperature dependent so radiation based PRMS algorithm has been applied to
vii
simulate snowpack melt in three seasons at a gap of two and half decades. It has partly impacted climate variability by reducing condensation-convection component of the energy balance.
Now the most important part of the snowmelt algorithm is snowline altitude and snow cover area. The latest technology and remote sensing techniques are employed to capture this information online, but the result and dependability are however limited for the operational hydrological models. Therefore, in this study, temperature and lapse rate dependent Varying Catchment Part (VCP) has been identified through the movement of its mean Saturated Snowline Altitude (SLA). It is then coupled with a monthly simulation model albeit the Applied Basic Oscillation (ABO) of the snowline. The VCP universal parameters, TLR of 0.554 °C/100 m and saturated snowline mean air temperature of 5.75 °C, were designed to estimate SLA using weekly mean temperature and ablating snowpack at and around Solang (2485 m altitude). The VCP model for SLA uses a nomogram at Solang while the ABO application extended to produce monthly snowline and Equilibrium Line Altitude (ELA). Its inter-annual/decadal variability gives a rate of rise in ELA of 11 m/yr, which is a possible signature of climate change in the region. The Snow Cover Area (SCA) has been determined through SLA, but the distributed hydrological model works with zonal SCA depletion curves. Therefore, combining the decay and Mathieu function formulations, the simulated parameters have been determined to evaluate zonal SCA depletion with depleting snow depth at base station. However, for SCA depletion in higher altitude zones, snow depth has to be augmented with proxy data.
Further, the actual snowmelt simulation has been used in order to formulate a better algorithm (temperature index, energy balance or mixed) irrespective of data
viii
constraints. Yet, simulation deficiency could not be resolved in case of above normal wind speeds. The algorithms in different watershed models for point snowmelt successfully simulate observed snowpack ablation (R2 up to 0.7 to 0.97) on weekly data. On the basis of performance criteria, the PRMS model results agree well with the full Energy Balance (EB) scheme on daily basis data, though SSARR is good on weekly basis. Finally, physical process based simple nomogram on temperature index has been developed in this thesis and it has been found that average condition nomogram is best suited for the varying climatic conditions. The glacier melt simulation needs more data than what is presently available. Although the relation between glacier melt and temperature may still be nonlinear yet the linear nomogram could be applicable for the temperature limits of 0.0 to 11.5 oC or more than 50% of snow cover area. The rest of the snow-free area is good for the rainfall-runoff model, which is beyond the scope of this study. Further, the performance of the nomogram has been compared with Martinec's SRM and HEC-HMS models by distributing the Beas basin (down to Manali Bridge) in three zones. The analysis of observed discharge data at Manali bridge and Palchan show reduced annual discharge and increased springtime discharge. The calibrated models have also been used to compute the variation in discharge for a unit degree rise in temperature.
TABLE OF CONTENTS
CERTIFICATE i
ACKNOWLEDGEMENTS iii
ABSTRACT v
CONTENTS ix
LIST OF FIGURES xiii
LIST OF TABLES xvii
LIST OF SYMBOLS AND NOTATIONS xix
CHAPTER 1 INTRODUCTION 1-25
1.1 GENERAL 1
1.2 NEED OF SNOW AND GLACIER MELT STUDY 2
1.3 THE STATE-OF-THE-ART 3
1.4 RESEARCH GAPS 4
1.5 OBJECTIVES OF THE STUDY 4
1.6 SCOPE OF WORK 5
1.7 STUDY AREA AND DATA 6
1.8 DISCHARGE DATA 19
1.9 ORGANIZATION OF THESIS 23
CHAPTER 2 REVIEW OF LITERATURE 27-62
2.1 GENERAL 27
2.2 TEMPERATURE AND SNOW VARIABILITY 29 2.3 SNOWLINE ALTITUDE AND SNOW COVER 39
2.4 SNOWMELT MODELING 47
2.5 SNOW AND GACIER MELT RUNOFF 54 ix
x
2.6 CONCLUDING REMARKS 59
CHAPTER 3 TEMPERATURE AND SNOW VARIABILITY 63-100
3.1 GENERAL 63
3.2 TEMPERATURE AND SNOW VARIABILITY 66
3.3 SPATIOTEMPORAL TEMPERATURE 66
3.3.1 Temporal Variability 69
3.3.2 Spatial Variability and Lapse Rate 71
3.3.3 Spatiotemporal Variability 76
3.3.4 Temperature Reconstruction 79
3.4 PRECIPITATION AND SNOW DEPTH 83 3.4.1 Precipitation with Temperature 85 3.4.2 Snow Depth with Temperature 89 3.4.3 Projection to Higher Elevation 92 3.4.4 Critical Temperatures and Snowpack Melt 93 3.4.5 Point Snowmelt on PRMS Model 97
3.5 CONCLUDING REMARKS 99
CHAPTER 4 SNOWLINE ALTITUDE AND SNOWCOVER 101-131
4.1 GENERAL 101
4.2 SNOW LINE ALTITUDE MODELLING 103 4.2.1 Development of VCP Model 104 4.2.2 Development of ABO Model 106 4.3 VCP AND ABO APPLICATION-VALIDATION 108 4.3.1 SLA Prediction by VCP 108 4.3.2 SLA and ELA Prediction by ABO 112
xi
4.3.3 Snowline on Melting and Disappearance: 114 4.4 EXTENDED APPLICATION OF ABO-VCP 117 4.4.1 ABO in Smoothening SLA 117 4.4.2 Change Detection on Shift in SLA 118 4.4.3 ELA and Global Estimates 121
4.4.4 Nomogram for Applications 122
4.5 SNOW COVER AREA DEPLETION MODEL 124 4.5.1 Modelling Snow Cover Area Depletion 124 4.5.2 Mathieu Function for Snow Cover
Depletion
127
4.6 CONCLUDING REMARKS 130
CHAPTER 5 INTERCOMPARISION SNOWMELT APPROACHES 133-158
5.1 GENERAL 133
5.2 POINT SNOW MELT SIMULATION 135 5.2.1 Temperature Index and Energy Balance 135 5.2.2 Algorithms in Watershed Models 136 5.2.3 Simulation Efficiency Criteria 142 5.2.4 Inter-comparison of Performance 143
5.2.5 Climate Variability Impact 148
5.2.6 Performance based Algorithms Selection 149 5.3 ENERGY BALANCE ON DAILY DATA 153
5.4 CONCLUDING REMARKS 157
CHAPTER 6 SNOW AND ICE MELT RUNOFF MODEL 159-177
6.1 GENERAL 159
6.2 TEMPERATURE INDEX MELT RUNOFF 161
xii
6.2.1 TtiMQ Nomogram 161
6.2.2 Nomogram Performance 169
6.2.3 Nomogram in Glacier Melt 170
6.3 SRM AND HEC-HMS MODELS 172
6.4 INTERCOMPARISON 173
6.5 CONCLUDING REMARKS 176
CHAPTER 7 SUMMARY AND CONCLUSIONS 179-189
7.1 Summary of Methodology 179
7.2 Important Findings 181
7.3 Limitation and Future Scope 188
REFERENCES 191
APPENDIX A POTENTIAL AND CRITICALTEMPERATURES 215 APPENDIX B EVALUATION CRITERIA AND MATHIEU
FUNCTION
217
APPENDIX C SNOWMELT ENERGY BALANCE ALGORITHMS 221 APPENDIX D SRM AND HEC-HMS MODELING 231 APPENDIX E UNCERTAINTIES MODELING BENEFITS AND
GLOBAL LINKAGE
253
PUBLICATIONS FROM THE THESIS 265
BIODATA 267
