Skip to main content

Error message

Workshop on SERVIR flood forecasting tools

Date 13–14 May 2019
Venue ICIMOD headquarters, Kathmandu, Nepal
Contact Persons
Type Workshop

Sustainable water resource management and reducing risks from flooding hazards are undeniably the utmost concerns of all countries of the Hindu Kush Himalayan (HKH) region. Confronted with limited knowledge and scarce resources, mountain-specific solutions to water-related problems have remained rather tenuous and generally ineffective. Most risk reduction and resilience strategies recommend proactive actions to minimize damage from adverse water events by anticipating such situations in advance.

Recognizing the need to enhance such capability in the region, the International Centre for Integrated Mountain Development’s (ICIMOD) SERVIR Hindu Kush Himalaya (SERVIR-HKH) Initiative, in collaboration with the SERVIR Applied Science Team (AST), has developed predictive tools to forecast high-impact weather events extending into sub-seasonal to seasonal forecasts, flash floods, and stream flows in river networks. Accordingly, ICIMOD is organizing a two-day workshop on SERVIR forecasting tools to bring together members of the SERVIR AST, SERVIR-HKH, flood forecasting professionals, and professionals working in flood forecasting and remediation from Bangladesh, Bhutan, and Nepal.

These tools provide information that is more reliable and actionable and show significant improvements over current deterministic forecasting systems. The tools are maintained in an open-access web portal for public assessment and feedback and to engage focal users in capacity development activities.

To be considered fit-for-purpose, systems that use a modeling approach need to undergo rigorous validation processes to ensure that the model simulates reality with an acceptable margin of error. The validation process also identifies areas for improvements to the model in terms of physical parameterization, boundary, and initial conditions. Thus, the workshop also includes an important session to validate model performance and verify forecasting skills to ensure the following:

  • SERVIR forecasting products are sustainable and deliver on the needs and expectations of user communities
  • Modeled results are evaluated against actual observations to measure predictive skill, spot errors, and test for significance within acceptable confidence level

Results from the disciplined validation method will inform decisions on whether to deploy as-is for operational setting, modify model configuration, or recalibrate and revalidate, with constructive feedback to SERVIR. Once forecast providers understand the strengths and weaknesses of a modeling system, they can issue warnings and advisories with confidence and communicate associated uncertainties in a coherent and actionable manner.


The objectives of this workshop are as follows:

  • Reorient participants to the statistics of model validation, and interpretation of comparative indices, error measures, and co-efficiencies
  • Familiarize participants on SERVIR applications and demonstrate their use in real-world situations
  • Develop skills and best practices through hands-on exercises in using forecasting applications to support decision making in sensible resource management and effectively dealing with emergency situations
Expected Outcomes
  • The participants are better informed on the products developed under SERVIR to support national river forecast service providers
  • Improvements in adoption and uptake of forecasting tools and application by relevant national partners for operational deployment
  • Communities at risk from flooding have access to improved, longer-lead flood warnings and advisories that are understandable and actionable