Akhtar et al. (2025) Integrated use of meteorological and hydrological indices for drought early warning in the mountainous catchments of the Himalaya-Karakorum-Hindukush region

Identification

• Journal: Scientific Reports
• Year: 2025
• Date: 2025-11-21
• Authors: Taimoor Akhtar, Haris Mushtaq, Muhammad Zia-ur-Rahman Hashmi
• DOI: 10.1038/s41598-025-25396-4

Research Groups

• School of Engineering, University of Guelph, Guelph, ON, Canada
• Energy, Climate Change and Just Transition Cluster, Deutsche Gesellschaft Für Internationale Zusammenarbeit (GIZ), Islamabad, Pakistan
• Environmental Science Center, Qatar University, Doha, Qatar

Short Summary

This study investigates the relationship between meteorological and hydrological droughts in the Upper Indus catchments of Pakistan using SPEI and SSI. It finds strong lagged cross-correlations between these indices in early Kharif months for Chenab, Jhelum, and Kabul catchments, which can be utilized for operational drought early warning and reservoir planning.

Objective

• To explore the relationship and coherence between meteorological and hydrological droughts in the Upper Indus catchments of Pakistan using the Standard Precipitation and Evaporation Index (SPEI) and the Standard Streamflow Index (SSI).
• To determine if the onset of hydrological droughts in these catchments can be predicted or characterized through meteorological drought indices, particularly considering seasonality.

Study Configuration

• Spatial Scale: Four key upper catchments of the Indus Basin in Pakistan: Upper Chenab, Upper Jhelum, Upper Indus, and Kabul. These catchments are located in the Himalaya-Karakorum-Hindukush region.
• Temporal Scale: Monthly data from January 1961 to December 2018 (58 years). Drought accumulation periods for SPEI ranged from 1 to 24 months, and for SSI from 1 to 12 months.

Methodology and Data

• Models used:
  • Standard Precipitation and Evaporation Index (SPEI) for meteorological droughts.
  • Standard Streamflow Index (SSI) for hydrological droughts.
  • Five probability distributions (log-logistic, Gamma, Pearson Type III, Lognormal, and GEV) were compared for SSI computation, with log-logistic selected as marginally better.
  • Cross-correlation and lagged cross-correlation analyses were performed to assess coherence between SPEI and SSI, incorporating seasonality.
  • R package 'SPEI' was used for index computations.
• Data sources:
  • Gridded CRU TS4.03 climate dataset (0.5° spatial resolution) for monthly precipitation and potential evapotranspiration (1901-2018).
  • Monthly observed streamflow data from 1961 to 2018 for the outlets of the four catchments: Chenab at Marala Barrage, Jhelum at Mangla Reservoir (upstream), Indus at Tarbela Reservoir (upstream), and Kabul at Nowshera gage.

Main Results

• Both SPEI and SSI effectively identified historical meteorological and hydrological drought events in the Upper Indus catchments, including the severe 1998–2002 drought.
• The log-logistic distribution was found to be marginally better for computing SSI values across the study sites.
• Meteorological droughts (SPEI) generally occurred with higher frequency but shorter durations compared to hydrological droughts (SSI), which were less frequent but longer-lasting.
• A strong mutual correlation was observed between drought duration and severity for both meteorological and hydrological droughts.
• Overall cross-correlations between SPEI and SSI were weak when seasonality was not considered.
• Strong monthly cross-correlations (r > 0.7) between SPEI and SSI were observed for Chenab, Jhelum, and Kabul catchments, particularly in late winter and early spring months (January-May for Chenab and Jhelum, March-August for Kabul).
• Significant lagged monthly cross-correlations (r > 0.7) were found between SPEI and SSI (with lags up to two months) for Chenab, Jhelum, and Kabul catchments during the early Kharif season (April to June).
• Poor cross-correlations between SPEI and SSI were observed for the Upper Indus catchment, primarily attributed to inaccuracies in gridded precipitation data in its upper regions.

Contributions

• This study is the first for the Indus Basin to systematically compare different probability distributions for Standard Streamflow Index (SSI) computation, recommending the log-logistic distribution.
• It provides a novel analysis of seasonal and lagged cross-correlations between meteorological (SPEI) and hydrological (SSI) drought indices in the complex, snow-dominated mountainous catchments of the Himalaya-Karakorum-Hindukush region, explicitly incorporating the impact of seasonality.
• Identifies specific lead times (up to two months) for early warning of hydrological droughts in the Chenab, Jhelum, and Kabul basins during the critical early Kharif season, offering direct applicability for operational drought monitoring, forecasting, and reservoir planning (e.g., Mangla reservoir) in Pakistan.
• Highlights data limitations in the Upper Indus catchment and suggests future research directions for improving drought monitoring in data-scarce regions.

Funding

The authors thank the Surface Water Hydrology (SWH) group at Water and Power Development Authority (WAPDA) for providing streamflow data. No specific projects, programs, or reference codes for funding were explicitly mentioned in the provided text.

Citation

@article{Akhtar2025Integrated,
  author = {Akhtar, Taimoor and Mushtaq, Haris and Hashmi, Muhammad Zia-ur-Rahman},
  title = {Integrated use of meteorological and hydrological indices for drought early warning in the mountainous catchments of the Himalaya-Karakorum-Hindukush region},
  journal = {Scientific Reports},
  year = {2025},
  doi = {10.1038/s41598-025-25396-4},
  url = {https://doi.org/10.1038/s41598-025-25396-4}
}