Visualizing the agricultural water cycle in irrigated paddy areas: Development of the drought projection methodology

Updated:December 1, 2020 (Tuesday)

-Effective tool for developing the climate change adaptation plan-

The National Agriculture and Food Research Organization (NARO) has developed a distributed water circulation model for large river-basin areas that incorporates reservoir operation and agricultural water movements. It simulates the basin-wide water cycles in spatial and temporal variations. Agricultural water movements are an essential aspect of understanding the hydrological system in Japanese river basin areas. The Distributed Water Circulation Model with Agricultural Water Use (DWCM-AgWU) is capable of predicting the river discharge in droughts. Its ability to make long-term projections under various conditions will support administrative organizations to develop water resource management plans applicable to the continuously changing environment under climate change.


In Japan, 70% of total water diversion from rivers is used predominantly for rice cultivation in irrigated paddy areas, and therefore the local agricultural water management has a direct impact on the river discharge. However, the repeated cycles of diversion from rivers, allocation of water to the irrigated paddies, and the return flow had been difficult to incorporate into the distributed water circulation model up to the present.

DWCM-AgWU developed by NARO is for simulating the river discharge during the irrigation period to predict droughts and offers a basis for adjusting local agricultural water management plans to adapt to the future environment under climate change. It is far more accurate than the conventional models by analyzing not only the natural hydrological cycles (precipitation, temperature, evapotranspiration, snowfall, snowmelt, river flow) but also the anthropological hydrological circulations (water use and management through dams, diversion weirs, and irrigation canals) to its analysis.

This dynamic distributed water circulation model can be utilized by various administrative organizations that are in charge of the agricultural water allocations to predict future risks of drought in rural areas, such as regional agricultural offices, the Japan Water Agency, and local governments.

DWCM-AgWU can also compute the probability of drought occurrence under climate change by feeding the precipitation and snowmelt projection data calculated from the climate change scenarios of major global climate models (GCMs). These projections are expected to contribute to principals and municipalities formulating local climate change adaptation plans, which is a legal responsibility prescribed in the Climate Change Adaptation Act since 2018.


Yoshida T., Masumoto T., KUDO R., TANIGUCHI T., and Horikawa N. (2012): Modeling of Basin‐wide Water Circulation Incorporating Water Allocation and Management Schemes in Large Irrigated Paddy Areas, Irrigation, Drainage and Rural Engineering Journal,DOI: 10.11408/jsidre.80.9.

Yoshida T., Masumoto T., Horikawa N., Kudo R., Minakawa H., and Nawa N. (2016): River basin scale analysis on the return ratio of diverted water from irrigated paddy areas,, Irrigation and Drainage,
DOI: 10.1002/ird.2040.

For Inquiries


Reference Information

Fig.1 (a) Overview of DWCM-AgWU.
(b) Comparison of the river discharge calculation accuracy between DWCM-AgWU and the conventional model.

(a) DWCM-AgWU shares the basic structure with conventional models for simulating the natural water circulation system. A river basin area is divided by grids, and each of the squares in 1km x 1km (mesh) represents the area's geographical attributes and river flows as in georama. The hydrological cycles of the river basin area are considered as the result of the interactions between these meshes. By feeding particular meteorological conditions to each of the meshes, the model simulates the interactions between the meshes to calculate the evapotranspiration, snowfall, snowmelt, and surface runoff after rains. The model also describes the river discharge increasing from upstream to downstream by simulating surface runoff joining a river flow by a mesh unit. Here, the river flow is directed according to the altitude gaps between meshes.

Based on the structure explained above, DWCM-AgWU is designed to incorporate the anthropogenic water allocations such as water discharge from dams and the diversion and return processes of agricultural use. After the user sets conditions to dams and the diversion weirs in the target area such as location, size, operation type, the purpose of the operation, the model calculates the volume of the water discharge from the dam and the total volume of diversion from the weirs based on the average river discharge data and facility operational rules. It is followed by the calculation of water movements of the irrigated water of being absorbed into the soil to form the underground water or to runoff to drainage canals after rainfalls. This series of computing processes realize the simulation of the entire hydrological circulation system of the river basin area.

(b) Calculated hydrograph compared to the observed hydrograph at the downstream station, Takada, located at the Seki River basin outlet. The calculated values with return flow using DWCM-AgWU (red line) is closer to the observed values than the calculated values without return flow using the conventional model (blue line). DWCM-AgWU proved its accurate computing ability, especially for the drought period.

Fig.2 (a) The whole view of the river-basin area, (b) Observed and calculated river discharge during the irrigation period.
DWCM-AgWU will project the drought risk and its scale at a particular location within the river-basin area. The calculated or estimated hydrograph for the diversion point i in River A shows that occasional water supply from reservoirs and dams upstream prevents the river discharge from going below the drought level. On the other hand, the calculated hydrograph for the diversion point ii in River B tributary shows the large fluctuation in river discharge, with several occurrences of drought.

Fig.3 The projections of river discharge in 2050~2070 compared to that in 1981~2000.
Long term projections using DWCM-AgWU used the precipitation projection data calculated from the climate change scenarios of major global climate models (GCMs).