Hydrological Modeling in the Upper Lancang-Mekong River Basin Using Global and Regional Gridded Meteorological Re-Analyses

Shixiao Zhang, Yang Lang, Furong Yang, Xinran Qiao, Xiuni Li, Yuefei Gu, Qi Yi, Lifeng Luo, Qingyun DuanWater 2023 · DOI

TL;DR

This study investigates how different meteorological re-analysis datasets (CFSR and CMADS) impact hydrological simulations in the upper Lancang-Mekong River Basin (LMRB). The findings reveal that while both datasets show similar spatial patterns, they lead to distinct hydrological responses, parameter sensitivities, and water balance components, with CFSR-driven simulations performing slightly better.

What Problem Does This Paper Address?

Accurate hydrological modeling is crucial for understanding water resources, predicting floods and droughts, and managing water in regions like the upper Lancang-Mekong River Basin (LMRB). However, a significant challenge in this region is the scarcity of ground-based meteorological stations, leading to uncertainties in hydrological simulations. To overcome this, researchers often turn to gridded meteorological re-analysis datasets. This paper addresses the critical question of how the choice of these re-analysis datasets (specifically, the global CFSR and the regional CMADS) influences hydrological model outputs and the resulting understanding of the basin's water cycle. It aims to clarify the strengths and weaknesses of these datasets in driving hydrological models.

Methods

The researchers employed the Soil and Water Assessment Tool Plus (SWAT+) model, a semi-distributed hydrological model, to simulate streamflow in the upper LMRB. They performed a comprehensive analysis involving:

  • Meteorological Data Evaluation: Six key meteorological variables (precipitation, maximum/minimum temperature, relative humidity, wind speed, and solar radiation) from CFSR and CMADS were compared using statistical indices like Pearson's correlation coefficient (R), percent bias (PBIAS), and root mean square error (RMSE).
  • Parameter Sensitivity Analysis: The Sobol method was used to identify which hydrological model parameters were most sensitive to streamflow simulation under each meteorological dataset.
  • Model Calibration and Validation: The SWAT+ model was calibrated and validated using observed streamflow data from the Chiang Sean hydrological station. The differences in calibrated parameter values between the two datasets were analyzed.
  • Water Balance Analysis: The simulated water balance components (PET, ET, deep recharge, percolation, surface runoff, and baseflow) were compared between the CFSR- and CMADS-driven simulations.

Datasets

The study utilized the following datasets:

  • Climate Forecast System Re-analysis (CFSR): A global, high-resolution meteorological re-analysis dataset covering the period 1979-2014.
  • China Meteorological Assimilation Driving Datasets for the Soil and Water Assessment Tool (CMADS) version 1.0: A regional re-analysis dataset for East Asia, used here for the period 2008-2013.
  • Digital Elevation Models (DEMs): 90m resolution DEMs from NASA and NIMA.
  • Land Use/Land Cover (LULC) Data: From the Earth Resources Observation and Science Center (USGS LULC 24 classification).
  • Soil Type Data: From the Food and Agriculture Organization of the United Nations (FAO).
  • Observed Streamflow Data: From the Chiang Sean hydrological station, provided by the Mekong River Commission.

The study focused on the common period of 2008-2013 for comparing CFSR and CMADS.

Key Insights and Ideas

  • Dataset Differences Matter: Even though CFSR and CMADS show similar spatial patterns for most meteorological variables, these subtle differences have a significant impact on hydrological model outputs. This highlights the importance of carefully selecting meteorological forcing data.
  • Parameter Sensitivity Shifts: The choice of meteorological dataset influences which hydrological parameters are identified as most sensitive and their optimal values. This suggests that a model calibrated with one dataset might not perform optimally with another.
  • Water Balance Divergence: The differing meteorological inputs lead to distinct simulated water balance components, particularly in precipitation conversion to streamflow and evapotranspiration. This implies that different datasets can lead to different interpretations of how water moves through the basin.
  • Regional vs. Global: While CMADS is a regional dataset, it's noted that regional datasets often assimilate more observations and resolve finer scales. However, the study found that the global CFSR dataset, despite its limitations, performed slightly better in this specific case.

Key Results

  • Meteorological Variable Comparison: CFSR generally showed higher precipitation, relative humidity, and wind speed, but lower temperatures and solar radiation compared to CMADS.
  • Statistical Performance: While both datasets showed strong correlations for temperature and solar radiation, precipitation and wind speed in CFSR exhibited lower correlations and higher biases compared to CMADS.
  • Hydrological Simulation Performance: The SWAT+ model driven by CFSR data resulted in slightly better streamflow simulation performance (higher NSE and R², lower PBIAS) during both calibration (2009-2010) and validation (2011-2013) periods compared to CMADS.
  • Water Balance Differences: CFSR-driven simulations showed higher precipitation but a lower conversion rate to streamflow (49% vs. 82% for CMADS), with higher evaporation and transpiration. CMADS-driven simulations showed higher percolation.

Limitations and Future Work

  • Limited Observation Data: The scarcity of ground-based meteorological stations and the use of only one streamflow gauging station (Chiang Sean) limit the ability to perform detailed bias correction and regional validation.
  • Short Common Period: The study used a common six-year period (2008-2013) for comparison, which might not capture all inter-annual variability.
  • Anthropogenic Activities: The study did not explicitly account for anthropogenic activities like dam construction, which significantly influence streamflow in the LMRB.
  • Climate Change: The short study period limited the ability to assess the impact of climate change on hydrological modeling.

The authors suggest future work could involve using climate forecast data for medium- to long-term streamflow prediction, analyzing how climate forecasts contribute to predictability, and potentially dividing the upper LMRB into several parts for more detailed regional analysis.

Summary updated

Abstract

Multisource meteorological re-analyses provide the most reliable forcing data for driving hydrological models to simulate streamflow. We aimed to assess different hydrological responses through hydrological modeling in the upper Lancang-Mekong River Basin (LMRB) using two gridded meteorological datasets, Climate Forecast System Re-analysis (CFSR) and the China Meteorological Assimilation Driving Datasets for the Soil and Water Assessment Tool (SWAT) model (CMADS). We selected the Pearson’s correlation coefficient (R), percent bias (PBIAS), and root mean square error (RMSE) indices to compare the six meteorological variables of the two datasets. The spatial distributions of the statistical indicators in CFSR and CMADS, namely, the R, PBIAS, and RMSE values, were different. Furthermore, the soil and water assessment tool plus (SWAT+) model was used to perform hydrological modeling based on CFSR and CMADS meteorological re-analyses in the upper LMRB. The different meteorological datasets resulted in significant differences in hydrological responses, reflected by variations in the sensitive parameters and their optimal values. The differences in the calibrated optimal values for the sensitive parameters led to differences in the simulated water balance components between the CFSR- and CMADS-based SWAT+ models. These findings could help improve the understanding of the strengths and weaknesses of different meteorological re-analysis datasets and their roles in hydrological modeling.
BibTeX
@inproceedings{zhang2023hydrological,
  author = {Shixiao Zhang and Yang Lang and Furong Yang and Xinran Qiao and Xiuni Li and Yuefei Gu and Qi Yi and Lifeng Luo and Qingyun Duan},
  title = {{Hydrological Modeling in the Upper Lancang-Mekong River Basin Using Global and Regional Gridded Meteorological Re-Analyses}},
  booktitle = {Water},
  year = {2023},
  doi = {10.3390/w15122209},
}
Figure