Swain, Janaki Ballav (2018) Streamflow Estimation in Ungauged Catchments through the Process of Regionalization. PhD thesis.
|PDF ((Full text is restricted up-to 15.09.2020)) |
Restricted to Repository staff only
Precise and dependable long-term streamflow record is a vital component in the modelling of hydrologic cycle. Quantitative and qualitative information regarding streamflow is needed for several practical application intended at various locations. However, lack of historical streamflow records makes its prediction in ungauged catchments a daunting task. Most of the catchments present across the globe are either partially gauged or completely ungauged. The scenario worsens in case of developing countries. The subject of ‘prediction in ungauged basins’ (PUB) is not prevalent in India. Regionalization is an alternative way for streamflow prediction in ungauged catchments. The present study deals with few regionalization approach for daily streamflow estimation in thirty-two identified catchments located in the Eastern and Southern India. The first category of regionalization i.e. hydrologic model-dependent approach is carried out using spatial proximity [Inverse Distance Weighted (IDW), Kriging and Global mean], regression analysis and physical similarity in conjunction with Soil and Water Assessment Tool (SWAT) for streamflow estimation in each catchment considered as ungauged in turn. The other type is hydrologic model-independent regionalization which is carried out using Flow Duration Curve (FDC). Area-Index, IDW, Kriging and regression analysis are the techniques used for information transfer from donor to receiver catchments. IDW and Kriging are the two techniques based on geographical distance between gauged and ungauged catchments produced superior results among all the implemented techniques in terms of Nash-Sutcliffe efficiency (NSE), Root Mean Square Error (RMSE)-observations standard deviation ratio (RSR) and Percent Bias (PBIAS). SWAT based regionalization techniques displayed better capability for capturing the variation in daily streamflow than regional FDC based approach. However, Information transfer between dissimilar catchments is likely to produce inaccurate results. Hence, linear [Principal Component Analysis (PCA) along with K-means clustering] and nonlinear [Self-Organizing Map (SOM) and Kernel PCA] classification techniques are executed to distribute the thirty-two catchments into few homogeneous groups based on the catchment attributes. SOM techniques performed best among all in classifying the catchments into four suitable groups. The potential impact of catchment classification on
streamflow regionalization is checked using the best two techniques i.e. IDW and Kriging on both classified and unclassified catchments. Half of the catchments exhibited more than 10% improvement in results when regionalization techniques are applied on classified catchments rather than unclassified ones. The overall outcome of the study proposes that application of suitable hydrological model, regionalization approach and catchment classification technique in an organized manner may produce reliable results in ungauged catchments with less uncertainty. Application of regionalization approaches considering the peripheral environment stationary may produce erroneous results. Thus, the impact of ever-changing Land use/Land cover and variable climate on streamflow regionalization is explored using IDW and Kriging technique in conjunction with SWAT. The outcomes suggest a substantial impact of these two factors on streamflow prediction. Hence, it is recommended to consider the non-stationary factors during the course of regionalization.
|Item Type:||Thesis (PhD)|
|Uncontrolled Keywords:||Ungauged catchment; regionalization; spatial proximity; regression; physical similarity; SWAT; FDC|
|Subjects:||Engineering and Technology > Civil Engineering > Water Resources Engineering|
Engineering and Technology > Civil Engineering > Construction Engineeing
|Divisions:||Engineering and Technology > Department of Civil Engineering|
|Deposited By:||IR Staff BPCL|
|Deposited On:||11 Sep 2018 11:16|
|Last Modified:||11 Sep 2018 11:16|
|Supervisor(s):||Patra, Kanhu Charan|
Repository Staff Only: item control page