Workshop on the Application of Hydrometeorological Seasonal Forecasts (THSF) Big Data for Water Management and Agriculture Support
Published June 14, 2025 | 10-minute read
ENG
THAI
Principles and Objectives
Forecasting is vital for planning and preparing for future situations, enhancing decision-making accuracy in hydrometeorology —the study of climate and water phenomena. Forecasting is essential for anticipating natural events and their potential impacts. In rapidly and severely changing climate conditions, the Seasonal Hydrometeorological Forecasting System project provides data with high spatial resolution of up to 0.01 degrees (approximately 1 kilometer), across Thailand, with forecasts ranging from 1 week to 6 months in advance. This local forecast data improves preparation for changing conditions and supports management of surface and groundwater resources for irrigation, matching plant needs at different growth stages. It enables better water usage planning and cultivation aligned with seasonal water availability, enhancing water resource management and agricultural operations.
This workshop on the application of real-time and forecasted hydrometeorological data is organized to demonstrate and present guidelines for utilizing the data. The main objective is to transfer knowledge to researchers who need hydrometeorological data for water management and agriculture (Technical Users) and end users from various agencies, to promote the application of data for decision-making in water and agricultural management under climate variability. Additionally, the workshop is an important opportunity to promote networking between researchers, academics, experts, and representatives from relevant agencies, especially government agencies responsible for national water and agricultural management planning. This will facilitate data exchange and pool knowledge and experience in using real-time and forecasted hydrometeorological data for effective decision-making, leading to sustainable development and adaptation to future climate conditions.
Training Workshop (Second meeting)
The workshop titled “Application of Big Data Hydrometeorological Seasonal Forecasts (THSF) for Water Management and Agricultural Support” was held at Room L422, 4th Floor, Arun Indrapalit Building, Irrigation Development Institute, Royal Irrigation Department, from June 5-6, 2025, between 8.30 AM to 4.00 PM. The primary objective of this research project was to demonstrate and showcase how to utilize the data. Its main goal was to transfer knowledge to Technical Users (researchers) and End Users (farmers) from various agencies who require hydro-meteorological data for water and agricultural management. It also aimed to promote the application of this data for decision-making in water and agricultural management under climate variability.
Furthermore, the workshop served as an important opportunity to foster networking and collaboration among researchers, academics, experts, and representatives from relevant agencies, especially government bodies responsible for national water and agricultural planning. This fostered the exchange of information, knowledge, and experience in using advance hydro-meteorological forecast data for effective decision-making, leading to sustainable development and adaptation to future climate conditions.
The workshop invited Dr. Somchai Baimuang, a meteorology expert; Dr. Sompop Sujarit, Advisor to the Bureau of Water and Hydrology Management; and Dr. Kanoksri Sarinnaphakorn, Head of Climate and Weather at the Hydro and Agro Informatics Institute, to be speakers. They provided valuable information, enhanced knowledge and skills, and facilitated effective collaborative networks. A total of 98 interested participants registered from various organizations, as listed below:
Organization | Number of Participants |
Rajamangala University of Technology Thanyaburi | 4 |
Department of Water Resources | 3 |
NECTEC | 4 |
Mahidol University | 4 |
Kasetsart University | 2 |
Chulalongkorn University | 4 |
Office of Agricultural Economics | 1 |
Panya Consultants Co., Ltd. | 3 |
Royal Irrigation Department | 9 |
Office of the National Water Resources (ONWR) | 5 |
Department of Royal Rainmaking and Agricultural Aviation | 6 |
Land Development Department | 4 |
Thai Meteorological Department | 3 |
East Water Public Company Limited | 3 |
Department of Groundwater Resources | 3 |
Department of Rice | 5 |
GISTDA | 1 |
Ministry of Natural Resources and Environment | 3 |
Fiscal Policy Office | 1 |
Department of Climate Change and Environment | 7 |
Hydro Informatics Institute (Public Organization) | 5 |
Metropolitan Waterworks Authority | 5 |
Agricultural Research Development Agency (Public Organization) | 2 |
National Disaster Warning Center | 1 |
Rubber Authority of Thailand | 2 |
SCG | 1 |
Department of Mineral Resources | 7 |
Total | 98 |
Figure 1: Pie chart summarizing the total participants by affiliated organization.
Figure 2: Group photo of workshop participants (both in-person and Zoom)
Figure 3: Mr. Chaiya Puangpholsap, Director of the Irrigation Development Institute, delivering the opening remarks.
Figure 4: Presentation on the Thailand Hydrometeorological Seasonal Forecasts (THSF) project, featuring high-resolution (1 kilometer) data for enhancing agricultural capabilities under climate variability challenges, by Asst. Prof. Dr. Natachet Tangdamrongsub, Research Project Leader.
Figure 5: Explanation of participant grouping for collaborative operations and analysis of trends and benefits of advance hydro-meteorological forecast variables by Asst. Prof. Dr. Saowanit Praprabnakorn, Project Researcher.
Seminar 1: Application of Geospatial Data and Advance Hydro-Meteorological Forecasting under Climate Variability by Dr. Somchai Baimuang
Dr. Somchai Baimuang, a distinguished expert in water management from the Agricultural Research Development Agency (ARDA) and the National Research Council of Thailand (NRCT), presented concepts and applications of geospatial data and advance forecasting to address climate change and variability. He emphasized the importance of understanding climate variability at both Temporal Scales (covering Weather (hours, days, months), Climate Variability (more than 1 month but not exceeding 10 years), and Climate Change (more than 10 years)) and Spatial Scales, including analysis of Thailand’s climate characteristics, atmospheric circulation patterns such as weather maps, monsoon troughs, and storms, as well as factors causing rainfall and floods in Thailand.
For applying meteorological data under varying climate conditions, Dr. Somchai provided examples of using advance weather forecasts combined with spatial, temporal, and weather characteristic data. Additionally, he discussed studies on meteorological factors during El Niño events using the Oceanic Niño Index (ONI) to assess the country’s water management situation. He also analyzed rainfall and temperature patterns in Thailand for the year 2024, comparing them with historical events using ENSO CODE data. Furthermore, he evaluated sweet corn yield using the Crop Growth Simulation Model (CGSM) combined with Remote Sensing (RS) and GIS techniques, and explored multi-dimensional drought studies including Meteorological Drought (e.g., SPI and Palmer Drought Index), Agricultural Drought, Hydrological Drought, and Socio-economic Drought.
Figure 6:Seminar 1 on the application of geospatial data and advance hydro-meteorological forecasting under climate variability by Dr. Somchai Baimuang.
Seminar 2: Application of Geospatial Data and Advance Hydro-Meteorological Forecasting for Agricultural Water Management and Disaster Prevention by Dr. Sompop Sujarit
Dr. Sompop Sujarit, Director of the Bureau of Water and Hydrology Management, presented on the development of forecasting technology and emphasized the importance of accurate and spatially detailed forecasts for agricultural water management and disaster preparedness. He pointed out that Thailand still has limitations in surface water storage, being able to store only about one-third of the total water volume. He noted that only two river basins, Ping and Mae Klong, can store sufficient water for demand, while other basins continue to face water shortages.
Dr. Sompop proposed the concept of a “Watershed Vulnerability Index” to assess and prioritize spatial water management planning. This index would reflect potential and risks from drought, floods, and climate change. He also stressed the importance of collaboration among agencies including government, local authorities, the private sector, and the public, in integrating and communicating data from multiple sources for sustainable water management.
Regarding reservoir management, Dr. Sompop indicated that it cannot rely solely on statistical models but requires dynamic analysis and water balance models to understand inflow patterns. Applying rainfall and runoff forecasts will help predict water inflow into reservoirs, and creating water balance models will make the assessment of inflow patterns more accurate. This data will be beneficial for decisions on water storage or discharge, managing drought and flood risks, seasonal agricultural water management, and assessing water security at local and national levels.
Figure 7: Seminar 2 on the application of geospatial data and advance hydro-meteorological forecasting for agricultural water management and disaster prevention by Dr. Sompop Sujarit.
Seminar 3: Application of Geospatial Data and Advance Hydro-Meteorological Forecasting in Water Management by Dr. Kanoksri Sarinnaphakorn
Dr. Kanoksri Sarinnaphakorn, Director of the Hydro and Agro Informatics Institute (Public Organization), presented guidelines for applying geospatial data and advance meteorological and hydrological forecasts for effective water management. She emphasized the importance of “Forecast Lead Time,” which is divided into: short-term (hourly-daily) for immediate warnings and daily water management; medium-term (weekly-monthly) for water allocation planning, such as crop planting decisions; and long-term (seasonal-yearly) for strategic planning and adaptation to climate change.
Dr. Kanoksri highlighted that hydro-meteorological forecast data, such as rainfall, temperature, wind, flow rate, and water levels, are crucial in all dimensions of water management. She stressed the need to translate “raw data” into “decision-making data” for practical use, such as flood assessment or adjusting water release plans. This will lead to the development of practical Early Warning Systems for both officials and the public. Examples of applications include Dashboards or Applications that provide timely access to forecast data and alerts, such as warnings for unsafe water levels to prepare for. The application of AI for knowledge communication was another approach Dr. Kanoksri presented through a Podcast that uses AI to convey information in an easily understandable way for the public.
While forecasting models face challenges in accuracy due to limitations of initial data and atmospheric variability, which represent “Uncertainty” that must be managed, Dr. Kanoksri proposed the concept of “Adaptive Planning” using Probabilistic Forecasts instead of relying on a single forecast value. This approach allows for flexible plans that can be adjusted to real-world situations. She also pointed out the role of new technologies such as AI/Machine Learning, Remote Sensing satellite data, and Big Data Analytics in enhancing accuracy and data accessibility, as well as the importance of cross-sector collaboration among government, academia, and local communities, and public participation in observing and reporting water situations to improve spatial analysis accuracy.
Figure 8: Seminar 3 on the application of geospatial data and advance hydro-meteorological forecasting in water management by Dr. Kanoksri Sarinnaphakorn.
Project Data Usage Introduction and Workshop
The project research team introduced the registration process and how to download high-resolution (1 kilometer) seasonal (6-month) forecast data covering all of Thailand through the project’s public website: https://alice-lab.com/thsf-th/. The main objective was to enable users to access data applicable in various fields, such as enhancing agricultural potential under climate variability, supporting decision-making for irrigation water management, drought index analysis, and other specific user needs, to effectively perform trend analysis and utilize variable data.
For tool application, the project research team presented methods for utilizing the project’s forecast data through processing on Google Earth Engine (GEE), along with a detailed explanation of GEE’s advantages and disadvantages. The session concluded with an introduction to basic Python usage and a demonstration of applying advance hydro-meteorological forecast data through processing using Python via Google Colab. This aimed to equip participants with the necessary knowledge and skills to maximize data utilization.
Figure 9: The project research team presented methods for utilizing the project’s forecast data through processing on Google Earth Engine (GEE) and Python via Google Colab.
To convey techniques for applying advance hydro-meteorological data in forecasting through Python and Google Earth Engine, workshop participants were divided into 6 groups based on their primary data application objectives:
Group 1 for academic benefit, Group 2 for disaster warning, Groups 3 and 4 for agriculture, and Groups 5 and 6 for water management. Each group processed data according to relevant variables and presented interesting results, as follows:
- Group 1 (Academic Benefit): Reducing Agricultural Water Use to Reduce Carbon: Forecasted the relationship between net shortwave/longwave radiation and net ecosystem exchange (NEE) to analyze the relationship between Leaf Area Index (LAI) and temperature, and net carbon balance. Results indicated high LAI values during September-December, which can be applied to reduce irrigation for carbon emission reduction.
- Group 2 (Disaster Warning): Flood Preparedness Planning in the Mae Klong Basin: Forecasted the relationship of rainfall, subsurface water, and soil moisture to analyze the soil permeability coefficient. It was found that heavy rainfall in July-August led to increased subsurface water in September-October, allowing for 1-2 months advance planning for flood response due to soil saturation.
- Group 3 (Agricultural Benefit): Analyzing Forecasted Rainfall Patterns in the Mun and Chao Phraya Basins: Forecasted the relationship of air temperature, evapotranspiration rate, and rainfall rate to analyze rainfall intensity and peak evapotranspiration periods. For the Chao Phraya Basin, a peak evapotranspiration trend was observed in July, followed by two decreasing periods: July-October and November-February. For the Mun Basin, evapotranspiration showed a decreasing trend from July-December, and rainfall increased in June, peaking in September-October. This data can be used for cultivation planning.
- Group 4 (Agricultural Benefit): Analyzing Optimal Rice Planting Period in the Chao Phraya Basin: Forecasted the relationship of Total Water Storage (TWS), Leaf Area Index (LAI), and mean soil moisture in the second layer (SoilMoist02) to analyze soil moisture suitable for rice cultivation. Results showed that September-November is the optimal period for rice growth due to high values of these variables.
- Group 5 (Water Management): Surface Runoff Forecasting in the Yom Basin: Applied rainfall, evapotranspiration, surface runoff, and Total Water Storage (TWS) to forecast surface runoff for assessing flood risk and planning planting/harvesting, as well as drainage. High surface runoff was found during September-November, which is beneficial for planning harvesting and draining standing water.
- Group 6 (Water Management): Forecasting Water Supply in the Chi Basin: Applied Total Water Storage (TWS), air temperature, and rainfall rate to forecast water supply for allocating water to various sectors. Results indicated that July-October had the highest rainfall and TWS values, suitable for reservoir storage for allocation from November onwards.
Figure 10: Atmosphere during the workshop: group processing of project variable data according to primary objectives and presentation of results by participants in each group.
Figure 11: Mr. Chawakorn Riewtrakulpaibul, Project Researcher, presenting certificates to workshop participants.
Participant Feedback
Participants considered the following hydro-meteorological forecast variables from the research project to be of utmost importance for water and agricultural management: Rainfall (1st priority, 14%), Surface runoff (2nd priority, 10%), and Subsurface runoff (1st priority, 8%).
Additionally, participants suggested adding other variables such as maximum and minimum temperature (Tmax, Tmin), atmospheric humidity, wind forecasts, and short- to long-range forecasts (3-7 days up to 3-5 years) for more comprehensive analysis. They also requested spatial data like DEM/Slope, Land Use/Land Cover (LULC) changes, soil and rock layers, and agriculture-related data such as crop calendars, crop types, land use, and vegetation indices (NDVI). Furthermore, there was a demand for data on water flow in rivers and reservoirs, water quality (e.g., turbidity), biodiversity rates, net radiation at the crop surface, and soil heat flux density to make the project’s data more complete and address a wider range of applications.
Figure 12: Chart summarizing participant feedback on hydro-meteorological forecast variables.
Figure 13: Chart summarizing participant feedback on the objective of using advance hydro-meteorological forecast data for monitoring and surveillance.
Summary of Knowledge and Understanding Survey and Workshop Evaluation
This workshop was highly successful in enhancing participants’ knowledge and understanding of applying geospatial data and advance hydro-meteorological forecasts. The percentage of participants with a high level of understanding increased from 27% to 64% after the training, demonstrating the effectiveness of the content and presentation format. Most participants recognized the immense benefits of the project’s high-resolution Big Data database for themselves and their agencies, especially in water management, agriculture, and drought warning. They also emphasized the necessity of forecast data and historical time-series data.
Forecasting drought-prone areas was identified as the primary objective for using forecast data (14%). Participants highly praised the practical presentations on using Python and Google Earth Engine (39% found it most beneficial), with over 90% considering these practical presentations extremely useful. The data dissemination channels through the project website and Google Earth Engine were also deemed appropriate and comprehensive.
Although knowledge clearly increased, participants still anticipated challenges in practical implementation, particularly regarding technical limitations (e.g., software, hardware, and personnel), integration with existing systems, and issues with data accessibility/sufficiency.
For future development, participants suggested increasing real-world case studies, extending practical training duration, and providing more foundational knowledge on data application. Additionally, they expressed a desire to learn advanced data analysis techniques, such as AI and Machine Learning, along with specialized applications in water resource management and integration with other data sources. The acquired knowledge will be used to enhance the capabilities of existing forecast data, develop new projects and applications, and transfer knowledge to colleagues and in teaching curricula. The research team also received recommendations to continue researching and synthesizing forecast data even after the project concludes to maximize benefits for the nation.
Additional Recommendations for Future Project and Activity Development
This project received excellent feedback, but for continuous development, the main recommendations are as follows:
- Activities and Training Content: The training duration should be extended, with a greater focus on using data in Python (including basics for beginners) and Google Earth Engine in more detail. The origin, data preparation, and processing guidelines should be explained more clearly. Python code should be improved to easily define time ranges and graph colors, and definitions or interpretation guidelines should be added for complex variables.
- Data and Platform: Data update frequency should be increased, and spatial data resolution should be finer, especially for disaster monitoring. Necessary variables should be added, and consideration should be given to linking data with actual hydro-meteorological models. The data analysis period should be extended to cover an entire year or multiple years to observe long-term impacts. Importantly, data should be shared via an API to allow daily 1-kilometer resolution forecast data and finer temporal data to be pulled.
- Support and Integration: The project should receive continuous budget support for sustainable development. There should be increased integration with government and private sector agencies. A system for data access for various user levels should be developed, and clear, detailed user manuals should be created. Additionally, research should be divided into specialized sub-groups, such as natural disasters, and the project should be promoted to the agricultural sector for utilization.
- General Suggestions and Problems Encountered: For future meetings, providing accommodation should be considered for 2-day events, and the number of in-person participants should be increased, as online participants experienced issues with audio and internet. The variety of lunch options should also be increased.
This project has developed a crucial hydro-meteorological forecasting model for water and agricultural management, focusing on providing high-resolution data (up to 1 kilometer) forecasted from 1 week to 1 season in advance. This aims to address the limitations of existing data, which often has low resolution and cannot provide advance forecasts. This model, operating on the NASA Land Information System (LIS) and installed on the NSTDA Supercomputer Center (ThaiSC), has been validated by comparing historical forecasts with reference data from ground-based stations and satellites such as SMAP, GRACE/GRACE-FO, and MODIS. The results demonstrate high accuracy across many variables nationwide, despite some limitations in certain areas (e.g., Southern and Western regions) and a decrease in accuracy with longer forecast periods.
From the knowledge transfer workshop, participants’ knowledge and understanding significantly increased. They clearly saw the immense benefits of this high-resolution Big Data database in water management, agriculture, and drought warning, especially the highly beneficial applications using Python and Google Earth Engine. However, participants also pointed out technical challenges, system integration, and data accessibility, and suggested continuous project support, longer model forecasting capabilities, and continuous data dissemination in the future to maximize benefits for farmers, water managers, and researchers in the country.
More information of the project please go to: https://alice-lab.com/thsf
HOMEPAGE
ALICE-LAB: Asian Land Information for Climate and Environmental Research Laboratory
Advancing Agriculture, Climate, Environmental, and Disaster Monitoring Through Satellite-Based Soil Moisture Retrieval and Applications in South East Asia
This workshop is focused on enhancing the understanding and utilization of satellite-based soil moisture data, particularly in the context of climate, agriculture, and disaster risk management. It aims to raise awareness among stakeholders in South East Asia about the significance of satellite-based soil moisture data, its applications, and the necessity of calibration and validation efforts to improve data accuracy.