Development of daily mean air temperature data with 50-m resolution for an information system identifying the suitable planting period for Yamadanishiki sake rice

Abstract

Degradation of the quality of Yamadanishiki sake rice has become more evident in the last 15 years, which is partly attributable to high temperature damage to the rice during heading. In response, an information system has been developed to determine the suitable planting period for this rice variety. A key input to this system is daily mean air temperature data at fine spatial resolution, and here an estimation method to calculate such data on a 50-m grid is presented. This database contains daily mean air temperature for 244 days in target regions from April through November over 15 years (1996)(1997)(1998)(1999)(2000)(2001)(2002)(2003)(2004)(2005)(2006)(2007)(2008)(2009)(2010). The database is obtained by applying a method whose essence is to convert air temperature to potential temperature and build estimation models of the potential temperature difference between an estimation site and existing observation site. The potential temperature differences are partitioned into two components: a "standard site component (Tssc)" and "estimation site component (Tesc)." Estimation models for Tssc and Tesc are developed using a radiative cooling scale, i.e., the difference of potential temperature between an upper level and ground level. Daily air temperature at each output grid was estimated to have a root mean square error of 0.5℃. Our method is able to acquire precise daily air temperature for past, present and future at the existing observation site, without newly costs for installation and maintenance of regular observing instruments.

Radiative cooling scale method for correcting hourly surface air temperature error in numerical weather prediction models

“…This difference in potential temperature is modelled in a linear regression equation using the RCS as a variable. RCS is a weather scale defined by Ueyama et al (2015) as the difference in potential temperature between an upper pressure surface and the ground surface. Ueyama (2013) computes the RCS using the NWP model output as follows:…”

“…This difference in potential temperature is modelled in a linear regression equation using the RCS as a variable. RCS is a weather scale defined by Ueyama et al (2015) as the difference in potential temperature between an upper pressure surface and the ground surface. Ueyama (2013) computes the RCS using the NWP model output as follows: $$$$ \mathrm{RCS}=\theta -{\theta}_G, $$$$ where θ is the potential temperature (in Kelvin) derived by the NWP model output at the upper level, and θ G is the observed potential temperature (in Kelvin) at ground level.…”

“…Previous research in this field has been reviewed by Frei (2014), Krähenman et al (2018), and Ueyama (2013). In our prior work, we developed a compilation technique for gridded daily SAT data with 50-m resolution using public data (Ueyama et al, 2015(Ueyama et al, , 2019, recognizing that gridded meteorological data provide the highest resolution for practical use in agriculture. We also developed a web application to distribute certain meteorological data, including SAT predictions based on public meteorological data, through an information communication device, for example, a smartphone or a tablet computer (Ueyama et al, 2018).…”

“…In this study, we develop a method to correct SAT data in NWP model outputs at an arbitrary site by applying RCS methods (Ueyama, 2013;Ueyama et al, 2015Ueyama et al, , 2018. The objective of the study was not to develop a method to improve NWP model output over large areas, but to propose a method to predict hourly SAT more precisely at specific geographical locations using public data, NWP model output, and observation data.…”