Water Resources Planning Branch, WRA, MOEAWater Resources Planning Branch, WRA, MOEA

Practical Guide to Hydrological Analysis and Review Essentials

Objectives of Hydrological Analysis and Definition of Design Discharge

• Estimation of design discharge: Mainly provides design discharges for hydraulic calculations and estimation of design water levels for flood control, drainage, river management, and river improvement planning.
• Disaster review and cause analysis: For flood events that have already occurred, the probability of occurrence (return period) is estimated to assess the causes and severity of the disaster and to provide a reference for subsequent planning and design.
• Return period criteria adopted: The frequency criterion shall be determined according to the purpose of the analysis. For example, river zone delineation generally adopts a 2-year return period; river improvement planning adopts larger return periods such as 25, 50, or 100 years; regional drainage planning adopts a 10-year return period and requires no overtopping for a 25-year return period.
• Definition of design discharge: The discharge at the outlet of a specific watershed or at an important control point within the watershed under a specified hydrological design standard (i.e., a specific return period).

Standard Operating Procedure for Hydrological Analysis

• Hydrological data collection and preprocessing: Collect hourly rainfall (or 10-minute rainfall) from meteorological and rainfall stations within the watershed, historical discharge records from water level/discharge stations, and spatial and physiographic data such as the Digital Elevation Model (DEM) of the watershed. Missing data are supplemented, and trend analysis, outlier detection, and anomaly detection are performed.
•  Rainfall and storm frequency analysis: Establish annual maximum rainfall series. Five theoretical probability distributions, including LN2, LN3, PT3, LPT3, and EV1, are fitted. Chi-square or K-S goodness-of-fit tests and error analyses are conducted to select the optimal probability distribution for estimating design storm rainfall for specific return periods.
• Establishment of design rainfall patterns: Common methods for establishing design rainfall patterns include the ranking method (rearranging observed rainfall), the alternating block method, and the SSGM method.
• Flood discharge analysis: Mainly includes the “design rainfall combined with rainfall-runoff model analysis method” (such as the Water Resources Agency dimensionless unit hydrograph method and triangular unit hydrograph method, which can derive a complete discharge hydrograph) and the “observed discharge frequency analysis method” (using annual maximum instantaneous discharge records, with at least 20 years of observation data).
• Comparison and selection of analysis results: Estimate return-period discharges at each control point and compare them with analysis results from previous literature or earlier reports to examine their reasonableness.

Key Points and Considerations for Hydrological Report Review

• Review of rainfall station selection:
  • Rainfall station selection should be based on clear and quantitative criteria (e.g., priority should be given to stations operated by the Central Weather Administration, Water Resources Agency, and Taiwan Power Company with records of 15–25 years or more, and with the latest year of record within the past 10 years).
  • If stations with shorter records or nearby stations outside the watershed are selected as representative stations, their correlation and reasons for selection must be explained.
• Review of storm rainfall frequency and duration:
  • Storm rainfall analysis should, in principle, be based on the average rainfall of the watershed (which may be calculated using the Thiessen polygon method or isohyetal method). Design durations should generally be 24 hours and 48 hours.
  • If hourly rainfall data are unavailable and one-day/two-day storm rainfall is converted using conversion coefficients, unreasonable results in which the converted 24-hour rainfall exceeds the 48-hour rainfall should be avoided.
•  Review of design rainfall patterns:
  • Verify whether the ranking method properly follows the Water Resources Agency’s commonly used principle of “placing the maximum-ranked percentage at the center, with larger values on the right and smaller values on the left in alternating order.”
  • Verify whether the time interval of the rainfall pattern (unit rainfall duration) is appropriately determined according to the watershed concentration time to avoid excessively large time intervals that may result in underestimated design discharges.
• Review of rainfall-runoff models:
  • Verify whether the basis for adopting infiltration losses (rainfall losses) has been reassessed, or whether the reasonableness of continuing to use values from previous studies has been evaluated.
  • Verify whether the method for calculating concentration time is reasonable and whether the adopted unit hydrograph has been simulated and validated using historical storm events.
• Review of discharge analysis results:
  • Verify whether outliers and anomalies have been detected before analyzing discharge data. If a discharge station is affected by upstream reservoir regulation or water diversion operations, it is not suitable for direct estimation of natural discharge.
  • Verify whether the discharge results show the abnormal condition of “upstream control-point discharge being greater than downstream discharge.”
  • When comparing with previous reports, investigate the causes of differences between the current and previous analyses separately in terms of “storm rainfall,” “rainfall pattern,” and “rainfall-runoff model.”

Practical Application Programs and System Tools

• Hydrological and Hydraulic Design Analysis System Platform: Developed by Professor Kuang-Tun Lee of National Taiwan Ocean University under commission from the Water Resources Planning Branch, based on the QGIS framework. It can rapidly and automatically delineate upstream watershed boundaries for control points, import project watersheds, and calculate key physiographic factors such as watershed area, river length, slope, and concentration time.
•  Classic Water Program: A practical tool developed by Chief Engineer Hsieh Cheng of the Sixth River Management Branch, Water Resources Agency. It includes hydrological, hydraulic, interpolation, and other modules and can perform batch calculations for one-day/two-day storm rainfall conversion coefficients, hourly rainfall format conversion, Thiessen polygon average rainfall and weight recalculation, frequency analysis plotting and goodness-of-fit testing, ranking-method rainfall pattern analysis, and peak discharge estimation using unit hydrographs.
• R programming language: Commonly used in practice for exploratory hydrological data analysis and preprocessing, including plotting time-series graphs, frequency plots, and box plots for detecting outliers and anomalies in discharge/rainfall data.

Development, Application, and Promotion of the SRH-ONE Two-Dimensional Hydraulic Model

Taiwan’s water environment has unique characteristics, including steep river slopes, rapid flows, and intense rainfall. As a result, during typhoon and flood periods, Taiwan is frequently affected by compound disasters such as flooding, debris flows, and landslides. With changes in the natural and social environments, river management and water resource applications are facing increasingly severe challenges. Therefore, numerical models are often required to provide quantitative analyses to support issue analysis and decision-making in hydraulic engineering.

To achieve the objectives of integrated improvement and adaptation of centrally administered river basins, including issues related to waterways, land flooding, blue-green network conservation, and waterfront connectivity, integrated watershed analysis numerical models must be developed to conduct quantitative and qualitative analyses for river management.

Building upon the technical research and development achievements of previous Taiwan–U.S. Appendix VIII cooperation projects, efforts are being made to develop and integrate the SRH (Sedimentation and River Hydraulics) series of numerical models and promote their application among relevant domestic organizations. This not only facilitates technology transfer resulting from Taiwan–U.S. cooperation in water resources engineering but also provides important scientific references for integrated watershed improvement and adaptation planning and river management decision-making.

Develop and integrate watershed analysis numerical models suitable for Taiwan’s water environment:

• Continue Taiwan–U.S. cooperation achievements and implement technology transfer: Continue the technical R&D achievements of previous Taiwan–U.S. Appendix 8 cooperation and implement technology transfer of the R&D results in accordance with the implementation schedule.
• Integrate multidimensional numerical models: Develop and integrate the SRH (Sedimentation and River Hydraulics) series of numerical models, including SRH-1D, SRH-2D, SRH-3D, U2RANS, as well as the watershed model SRH-W and estuarine model SRH-Coast currently under development. These models must be applicable to Taiwan’s unique water environment characteristics, such as steep river slopes, rapid flows, and intense rainfall.

Enhance river management and decision analysis capabilities:

• Support issue analysis and decision management: Develop integrated watershed analysis numerical models to provide quantitative analyses supporting issue analysis and decision management in hydraulic engineering and to address compound disasters such as flooding, debris flows, and landslides.
• Support integrated watershed improvement and adaptation planning: Provide quantitative and qualitative analyses for integrated improvement and adaptation planning of centrally administered river basins, including waterways, land flooding, blue-green network conservation, and waterfront connectivity, to support river management decision-making.

Establish a user-friendly operating interface and comprehensive user manuals:

• Develop an integrated operating interface (SRH-ONE): Initially develop an operating interface for integrated watershed analysis numerical models, simplify the establishment of the SRH model operating environment, improve user interfaces for multiple models, and address difficulties encountered by hydraulic engineering personnel in learning to operate multidimensional models.
• Compile Chinese-language operating manuals and case studies: Compile Chinese-language operating manuals and application cases for the SRH-2D hydraulic and movable-bed models, as well as SRH-1D and SRH-3D, to provide localized teaching demonstrations and enhance user learning effectiveness.

Promote numerical model applications and provide technical consultation:

• Expand model applications and user groups: Promote numerical model applications among relevant domestic organizations, such as the Water Resources Agency, its subordinate agencies, consulting companies, research institutions, colleges, and universities, expanding from primarily SRH-2D to SRH-1D, SRH-3D, and U2RANS.
• Provide professional training and technical consultation: Through professional training for seed trainees, issue-oriented model analysis, training courses featuring feedback and exchange on application results, and technical consultation services, strengthen numerical model functions and expand their scope of application.

Analysis and Application of Hydrological Scenarios for Flood Management under Climate Change

Downscaled rainfall data provided by TCCIP are applied to analyze changes in rainfall under climate change for centrally administered river basins throughout Taiwan and to establish a framework for flood-control hydrological scenario analysis under climate change in Taiwan. The purpose is to assess the impacts of climate change on flood-control safety in centrally administered river basins and provide a reference for climate change adaptation objectives in watershed adaptation planning.

Collection, compilation, and analysis of international cases:

• Explain climate change scenarios and fixed global warming conditions.
• Examine Taiwan’s climate change hydrological scenario data and recommendations for scenario settings.
• Assess meteorological and hydrological impacts under IPCC AR6 scenarios.
• Collect and analyze national- or global-scale impact assessment reports to understand international results and methodologies for assessing climate change impacts on extreme rainfall and flood discharge, including reports from the United States, European Environment Agency, New Zealand Ministry for the Environment, Environment and Climate Change Canada, and the Intergovernmental Panel on Climate Change (IPCC).

Analysis of climate change-induced rainfall changes in centrally administered river basins:

• Analyze climate change-induced rainfall changes for 26 centrally administered rivers, including basic information such as main-channel length, watershed area, and counties/cities through which they flow.
• Present the results of dynamically downscaled extreme rainfall impact analyses.
• Present the results of statistically downscaled extreme rainfall impact analyses.
• Compare analysis results from different downscaling products.
• Examine the characteristics of short-duration extreme rainfall impacts.
• The appendix also provides dynamically and statistically downscaled extreme rainfall analysis results for the 26 centrally administered river basins.

Methods and framework for analyzing downscaled rainfall and discharge under climate change:

• Develop a framework and process for flood-control hydrological analysis under climate change. The framework is mainly divided into two parts:
  • Extreme rainfall analysis under climate change: Includes grid matching for watershed control points, baseline and future extreme rainfall frequency analyses, and calculation of scenario-based extreme rainfall increments. Statistical downscaling data are recommended for examining extreme rainfall characteristics because they can more reasonably reflect the uncertainties of climate change.
  • Flood discharge analysis.
• In the future, hydraulic calculations may be conducted in conjunction with storm-surge level data for different return periods at river mouths developed by the National Science and Technology Center for Disaster Reduction (NCDR) to assess the impacts of climate change on river-channel risks.

Climate Change – Flood Hydrology Scenarios and Application

To evaluate the impact of climate change on flood safety within central government-administered river basins, the Taiwan Climate Change Projection and Information Platform Project (TCCIP) downscaled rainfall data has been applied. This data serves as the foundation to establish a framework for climate change flood scenario analysis in Taiwan, providing essential reference for basin-wide adaptation planning and climate resilience strategies.

 International Reference and Scenario Analysis

• Description of climate change scenarios and fixed-temperature pathways.
• Recommendations for the use and definition of Taiwan-specific hydrological climate scenarios.
• Impact evaluation of IPCC AR6-based climate and hydrological changes.
• Compilation and analysis of global and national-level impact assessment reports, including those from:
  • The United States, European Environment Agency, Ministry for the Environment of New Zealand, Environment and Climate Change Canada, and the IPCC.

 Rainfall Change Analysis for Central Government River Basins

• Analysis of rainfall change for 26 centrally managed rivers, considering:
  • Mainstream length, watershed area, administrative districts.
• Presentation of results from:
  • Dynamically downscaled rainfall impact analysis
  • Statistically downscaled rainfall impact analysis
  • Comparative study of different downscaling methods
  • Characteristics of short-duration heavy rainfall events
• The appendix includes dynamic and statistical rainfall analyses for all 26 river basins.

Framework and Case Study of Downscaled Rainfall & Flow Analysis

Flood Hydrology Analysis Framework under Climate Change:

• The framework consists of two major parts:

1. Rainstorm analysis
   • Includes grid-mapping to control points, baseline and future frequency analysis, and scenario-based rainfall increment calculations.
   • Use of statistical downscaling is recommended for better representation of climate uncertainties.
2. Flood flow analysis
   • Case studies are conducted to assess climate change impacts on design rainfall and flood discharges.

• Additionally, the analysis recommends integrating storm surge data from NCDR (with varying return periods) for riverine hydraulic modeling to assess climate-induced flood risks near river mouths.

Figure 1: Impact analysis of statistically downscaled rainfall at river estuary

Figure 2: Two-day rainfall increase percentage for Taiwan’s four major regions

Development and Promotion of SRH-One 2D Hydrodynamic Model

Taiwan’s water environments are characterized by steep river gradients, rapid flows, and intense rainfall. These factors often lead to compound disasters such as flooding, landslides, and debris flows during typhoons or storm events.

As environmental and societal conditions continue to evolve, river management and water resource planning face increasing complexity. Therefore, numerical models are vital tools to support engineering analysis and decision-making.

To achieve the goal of comprehensive improvement and adaptation for centrally managed basins, including topics like channel realignment, floodplain management, green-blue network conservation, and waterfront integration, it is necessary to develop integrated watershed analysis models capable of both quantitative and qualitative assessment.

Model Development Goals

1. Build localized watershed models suited to Taiwan's water environment
   • Continue outcomes from Taiwan–US collaboration (Appendix VIII) to ensure effective technology transfer and application.
2. Integrate multidimensional SRH modeling systems
   • Combine SRH (Sedimentation and River Hydraulics) model series:
     • SRH-1D, SRH-2D, SRH-3D, U2RANS, and in-development modules: SRH-W (watershed) and SRH-Coast (estuary).
   • All models are adapted for Taiwan’s unique conditions (steep slopes, high flow velocity, extreme rainfall).

 Enhancing River Management and Decision-Making

• Model-aided engineering analysis and decision support
  • Numerical tools assist in quantifying the impacts of complex hazards and improving engineering strategy.
• Support integrated river basin improvement and adaptation planning
  • Assist in quantitative/qualitative evaluations for policy and decision-making across various planning themes.

User-Friendly Interfaces and Documentation

• Development of SRH-One unified interface
  • Simplifies SRH model operations and supports a variety of user experiences for engineers unfamiliar with multidimensional models.
• Localized user manuals and case studies
  • Includes Chinese manuals and practical examples for SRH-1D, SRH-2D, and SRH-3D to enhance learning and application.

 Promotion and Technical Support

• Expanding user base and model application
  • Promote model usage among government agencies, consultants, researchers, and universities.
  • Expand from SRH-2D to other modules (SRH-1D, SRH-3D, U2RANS).
• Professional training and technical consulting
  • Conduct seed-user training, scenario-based model analysis, feedback integration, and application workshops.
  • Provide ongoing technical advisory services.

Figure 3: SRH model promotion and usage strategy

Figure 4: Cross-domain module development for SRH model system