Ray et al. (2026) Flow Regime Shifts in Eastern India Under Changing Climate: A Causality and Trend Perspective

Identification

• Journal: Earth Systems and Environment
• Year: 2026
• Date: 2026-01-06
• Authors: Soubhagya Laxmi Ray, AMBIKA PRASAD SAHU, J. C. Paul, Sanjay Kumar Raul, Prachi Pratyasha Jena, Santosh S. Palmate, Sonam Sandeep Dash
• DOI: 10.1007/s41748-025-00996-2

Research Groups

• Department of Soil and Water Conservation Engineering, College of Agricultural Engineering and Technology, Odisha University of Agriculture and Technology, Bhubaneswar, Odisha, India
• Department of Biological and Agricultural Engineering, Texas A & M University, Texas, TX, USA
• Physical Science and Engineering Division, King Abdullah University of Science and Technology, Thuwal, Saudi Arabia

Short Summary

This study integrates multi-site trend analysis of rainfall, maximum temperature, minimum temperature, and streamflow to unravel the complex links between meteorological forcings and anthropogenic disturbances on the Mahanadi River Basin's streamflow regime. It reveals heterogeneous, site-specific alterations in streamflow, primarily driven by declining rainfall and rising temperatures, with anthropogenic factors significantly influencing lower basin responses.

Objective

• Perform a comprehensive long-term trend and change point analysis to map climate variability (Rainfall, Tmax, Tmin) and hydrological response across the Mahanadi River Basin (MRB).
• Quantify the magnitude of regime shifts in flow quantiles following identified change points.
• Investigate the causal interlinkages between meteorological variables and streamflow dynamics to identify sensitive drivers.

Study Configuration

• Spatial Scale: Mahanadi River Basin (MRB), India, with a total drainage area of 141,589 km². The study utilized data from 16 gauging stations distributed across the upper, middle, and lower sub-basins.
• Temporal Scale:
  • Rainfall data: 1901–2022 (122 years).
  • Maximum and minimum temperature data: 1951–2022 (72 years).
  • Streamflow data: 1980–2018 (39 years).
  • Land Use/Land Cover (LULC) data: Decadal assessments for 1985, 1995, 2005, and 2015.

Methodology and Data

• Models used: Mann-Kendall (MK) test, Modified Mann-Kendall (MMK) test, Pettitt’s test, Theil-Sen’s Slope Estimator, Sequential Mann-Kendall (SQ-MK) test, Vector Autoregression (VAR) framework, Granger Causality, Impulse Response Functions (IRFs), Augmented Dickey-Fuller (ADF) test, Akaike Information Criterion (AIC), Weibull plotting position formula (for Flow Duration Curves).
• Data sources:
  • Daily gridded rainfall (0.25° × 0.25°), maximum temperature (1° × 1°), and minimum temperature (1° × 1°) from the India Meteorological Department (IMD), Pune.
  • Daily observed streamflow from the Central Water Commission (CWC), Bhubaneswar, for 16 gauging stations.
  • Decadal Land Use/Land Cover (LULC) classification information from Roy et al. (2016) (100 m resolution) for 1985, 1995, 2005, and Moderate Resolution Imaging Spectroradiometer (MODIS) data (resampled to 100 m) for 2015.

Main Results

• Rainfall Trends: A majority of stations (54.73%) in the MRB show statistically significant decreasing trends in annual rainfall, with similar dominant decreasing trends observed for monsoon (51.74%) and pre-monsoon (56.22%) rainfall. Localized increases were identified in the southern and southeastern Lower MRB, with change points mostly confined to the early 1980s.
• Temperature Trends: Maximum temperature (Tmax) exhibits a strong and widespread increasing trend, with over 90% of stations showing statistically significant annual increases, most pronounced during the post-monsoon season. Minimum temperature (Tmin) trends are varied, with significant decreasing trends more prevalent in pre-monsoon (47% of stations) and monsoon (38% of stations) periods, suggesting an increasing diurnal temperature range.
• Streamflow Trends: Streamflow shows a mixed response across the basin. An annual decreasing trend was observed at Andhiyar Khore (Upper MRB), while the Kesinga station (Lower MRB) shows a significant increasing trend. Significant pre-monsoon streamflow decreases were observed at 43% of stations, particularly in the upper and middle MRB. Conversely, Kantamal and Kesinga stations consistently demonstrate strong increasing trends across multiple seasons.
• Flow Regime Shifts: Post-change point analysis reveals a predominant pattern of decreased streamflow across high (Q10), medium (Q50), and low (Q90) flow conditions at most stations in the upper and middle basins (e.g., Jondhra: Q10 -64.05%, Q50 -56.72%; Andhiyar Khore: Q90 -66.68%). In contrast, Kantamal and Kesinga in the lower basin exhibited remarkable increases across all flow quantiles (e.g., Kesinga: Q90 +109.34%; Kantamal: Q90 +92.34%), plausibly reflecting increasing rainfall trends in the Lower MRB, amplified runoff due to LULC changes, or managed releases from the Hirakud reservoir.
• Causality and Anthropogenic Influence: Granger causality analysis confirms rainfall as the dominant predictive driver of streamflow (statistically significant at 87.5% of stations). The direct predictive link from temperature to streamflow is weak and localized. Anthropogenic landscape transformations, including a notable expansion of agriculture (from 57.7% to 60.9%) and a near tripling of built-up areas (from 1.34% to 3.91%) between 1985 and 2015, contribute to the spatial heterogeneity in streamflow responses and modulate climatic drivers, especially in the lower basin where dam operations and water abstractions play a significant role.

Contributions

This study's primary novelty lies in its methodological integration at a comprehensive spatial scale for improved temporal characterization of the basin’s hydrological shifts. It is the first study in the Mahanadi River Basin to: 1. Combine non-parametric trend/change-point analysis with a robust econometric causality framework (Vector Autoregression, Granger Causality, and Impulse Response Functions). 2. Apply this integrated framework across 16 spatially distributed gauging stations to capture the heterogeneous basin-wide response, rather than focusing on a single sub-basin. 3. Quantitatively link observed climate drivers to specific, quantified shifts in the complete flow regime (Q10, Q50, Q90) before and after identified change points, providing direct evidence for targeted, region-specific water management.

Funding

The authors did not receive support from any organization for the submitted work.

Citation

@article{Ray2026Flow,
  author = {Ray, Soubhagya Laxmi and SAHU, AMBIKA PRASAD and Paul, J. C. and Raul, Sanjay Kumar and Jena, Prachi Pratyasha and Palmate, Santosh S. and Dash, Sonam Sandeep},
  title = {Flow Regime Shifts in Eastern India Under Changing Climate: A Causality and Trend Perspective},
  journal = {Earth Systems and Environment},
  year = {2026},
  doi = {10.1007/s41748-025-00996-2},
  url = {https://doi.org/10.1007/s41748-025-00996-2}
}