Japanese
Home > Services > Climate Risk Management > Pilot Project in Agriculture

Pilot Project to Develop Climate Risk Management Techniques for Agriculture

Introduction

Paddy rice is a staple food of the Asian region. As production volumes rise worldwide, competition in breed improvement and cultivation technology development becomes increasingly more intense. Meanwhile, although production volume has decreased due to population decline in Japan, demand for higher quality has gradually risen. Recent years have frequently seen extremely hot summers in association with global warming and decadal climate variability, and this influences the quality of rice with issues such poor grain filling.

The Japan Meteorological Agency (JMA) and the National Agricultural and Food Research Organization (NARO) have started a pilot project to develop an early warning system using two weeks forecasts. JMA also engages in dialogue with local agricultural organizations toward the development of a User Interface Platform (UIP).

The Rice Early Warning System

Background

Various studies have been conducted on the influence of anomalous climate conditions (Table 1). Based on the results obtained, climate prediction information is considered to have the potential to reduce the impact of adverse climate events.

Table 1 Alarm trigger temperatures and rice crop risk/action (NARO Tohoku Agricultural Research Center)

Period

 Alarm trigger Risk Required action

Around July

 20°C or below Sterility Water management

Around August

 27°C or above Poor grain filling Growth management

Procedure

JMA experimentally provides temperature prediction data with a 1-km resolution based on its two-week forecasts and NARO's normal dataset (Figure 1). The procedure is as follows:

  • JMA operationally issues two-week forecasts twice a week and conducts related calibration using model output statistics (MOS) methods.
  • NARO maintains a normal temperature dataset with a 1-km resolution (Kanno,1997).
  • Temperature data over a 1-km mesh covering the period up to two weeks ahead are obtained from a combination of two-week forecasts and the normal temperature dataset.
  • System users receive early warning information relating to the areas for which they are registered (Figure 2).

Figure 1

Figure 1 Procedure for 1-km mesh temperature data

Figure 2

Figure 2 Experimental provision to users

Verification

  • The reliability of forecast probability is relatively high for cold/hot conditions (Figure 3).
  • Users can take measures based on their Cost/Loss ratio.

Figure 3

Figure 3 Verification of second-week temperatures
Season: mid-July to early-August (left), August (right)
Region: 17 weather stations in the Tohoku district.


User Interface Platform Development

JMA engages in ongoing dialogue with local agricultural organizations to promote the use of climate information in agricultural decision making. As local-government agricultural staff are usually in close contact with farmers in their areas, dispatching climate information to agricultural organizations supports UIP development. One such organization has performed investigation using one-month forecast data to predict the ideal timing for rice harvesting (Figure 4), and will provide agricultural advice to farmers based on the results in 2014. This is a best practice for UIP development.

Figure 4

Figure 4 User Interface Platform (UIP) development