Prediction of climatic suitability for wine grape production under the climatic change in Hokkaido

Transcription

1 Short Paper Journal of Agricultural Meteorology 72 (3-4): , 2016 Prediction of climatic suitability for wine grape production under the climatic change in Hokkaido Manabu NEMOTO a,, Tomoyoshi HIROTA a and Tomonori SATO b a National Agriculture and Food Research Organization (NARO), Hokkaido Agricultural Research Center (HARC), Hitsujigaoka 1, Toyohira, Sapporo, Hokkaido , Japan b Faculty of Environmental Earth Science, Hokkaido University, Kita-10, Nishi-5, Sapporo, Hokkaido , Japan Abstract This paper investigates the shifts in climatic suitability for cultivating grapes (Vitis vinifera) in Hokkaido, Japan, throughout the 20th and 21st century. To assess area suitability over time, we used two climate model datasets and three climatic criteria (summer temperature, summer precipitation, and winter temperature). The optimum area for growth gradually enlarged from the 1990s to present (2010s), and this area will cover almost all of the Hokkaido region by the 2050s. Even when considering model uncertainties, almost all low elevation areas in Hokkaido are projected to be suitable for V. vinifera cultivation in the future. Key words: Climate change, Global warming, Vitis vinifera, Wine grapes. 1. Introduction The main places for the production of table grapes in Japan are the Yamanashi, Nagano, and Yamagata prefectures, and the grape yield for these three prefectures constituted 50% of the total grape harvest in Japan in 2014 (Ministry of Agriculture, Forestry and Fisheries, 2015). In contrast, the grape harvest for Hokkaido, the northernmost prefecture in Japan, only made up approximately 4% of the total harvest in 2014 (Ministry of Agriculture, Forestry and Fisheries, 2015). The main reason for this difference is thought to be climate conditions. Hokkaido has long cold winters and heavy snow, and these climate features are disadvantageous for fruit production. Interestingly, the number of wineries in Hokkaido has increased from 6 in 1999 to 22 in 2014 (Nagata et al., 2014). Although cold-tolerant German varieties of grapes, for instance, Kerner, Zweigeltrebe, and Müller-Thurgau, have been planted in Hokkaido for many years, many wineries are nowadays cultivating more varieties requiring high temperatures, for instance, Pinot Noir, Chardonnay, and Merlot. Our main interest is to determine whether climate conditions in Hokkaido at present and in the future will be suitable or not for the cultivation of European grape varieties for wine, specifically, Vitis vinifera. In this paper, we describe shifts in the climatic suitability for V. vinifera cultivation in Hokkaido, Japan, from the 20th to the end of the 21st century based on data obtained from climate change models. To assess climatic suitability for cultivating V. vinifera in Hokkaido from the 20th to 21st century, we selected the main places for wine grape production (Fig. 1) and used the following two climate model datasets to examine changes in the spatial distribution of three climatic criteria, namely, summer temperature, summer precipitation, and winter temperature. The first climate model dataset (MIROC 3.2 Hires, which we will describe in section 2.1) provides long-term data from 1900 to 2100 with high spatial resolutions of 1 km. The other dataset has nine outputs from multiplication of three general circulation models (GCMs) including the first dataset and three regional atmospheric models (RAMs) (RECCA, which we will describe in section 2.2) for two decadal periods, which include the present (1990s) and the future when the global mean surface air temperature will have increased by approximately 2 relative to the present (1990s) climate. The latter one is based on the A1B scenario. Therefore, our results considered the uncertainty of models based on the A1B scenario. 2.1 Long-term climate predictions We used the 1-km spatial resolution interpolated data from the high resolution version of the Model for Interdisciplinary Research on Climate (MIROC 3.2 Hires; Hasumi and Emori, 2004) and data that were based on scenario A1B found in the Special 2. Data and Methods Received; December 2, Accepted; April 25, Corresponding Author: nemo@affrc.go.jp DOI: /agrmet.D Fig. 1. Main places of grape production for wine in Hokkaido

2 Journal of Agricultural Meteorology 72 (3-4), 2016 Report on Emissions Scenarios (SRES; Nakicenovic and Swart, 2000) for climate predictions. The daily output data were interpolated into a 1-km mesh grid by Okada et al. (2009). The daily values of every 10-y mean throughout the 20th and 21st century were used to investigate the long-term variations in climate conditions relevant to the cultivation of V. vinifera. Bias correction was conducted by reference to the linear scaling approach (Teutschbein and Seibert, 2012) while using climate mesh data ( ) of the Japan Meteorological Agency (JMA). 2.2 Consideration of uncertainty due to the models We further compared spatial distributions of the climatic suitability for wine grape production in Hokkaido among different models. The inconsistency between the models represents the uncertainty due to the model predictions. A dataset developed by the Research Program on Climate Change Adaptation (RECCA; Inatsu et al., 2015) was used to assess this uncertainty. This daily dataset is composed of two decades of data for present and future conditions, which corresponded to the 1990s and decades when the global mean surface air temperature will have increased by approximately 2 relative to the present climate in each GCM under the SRES A1B scenario, respectively. Those were simulated by nine dynamical downscaling (DDS) experiments with combinations of three GCMs and three RAMs. The three GCMs used included (i) MIROC 3.2 Hires, (ii) MPI (fifth- Fig. 2. Long-term changes in climatic suitability for grape cultivation of Vitis vinifera in Hokkaido. a) The mesh grid was judged to to be suitable for wine grape production where all three criteria were satisfied. b) Summer temperature: white area shows the unsuitable area due to temperatures lower than c) Summer precipitation: black area shows the unsuitable area due to precipitation in excess of 1200 mm. d) Winter temperature: black area shows the unsuitable area due to temperatures lower than

3 M. Nemoto et al.:climatic suitability for wine grape production in Hokkaido generation atmospheric GCM of the Max-Plank-Institut fur Meteorologie), and (iii) NCAR (Community Climate System Model version 3 of the National Center for Atmospheric Research). The RAMs used included (i) NHM (Non-Hydrostatic Model developed by JMA and the Meteorological Research Institute), (ii) WRF (Advanced Research Weather Research and Forecasting Model version 3.2.1), and (iii) RSM (Regional Spectral Model developed by Scripps Experimental Climate Prediction Center). All DDS results were corrected for bias by offsetting the difference of the present climatology between observations and DDS results. Spatial resolution of this data was approximately 10 km. For more details, please refer to Inatsu et al. (2015). The future decade for the RECCA dataset corresponds to the DDS of the MIROC run targeted at the 2050s. 2.3 Judging climatic suitability for grape cultivation Climatic suitability on a mesh grid was judged against the following three criteria after Hannah et al. (2013). These criteria included the summer temperature, summer precipitation, and winter temperature. The summer temperature was defined as the average temperature during the growing season, where the growing season extends from April to October. The mesh grid was judged to be suitable for cultivation of all varieties of V. vinifera when the summer temperature was in the optimum range from 13.1 to 20.9 (Hannah et al., 2013). The summer precipitation was defined as the total amount of precipitation during the growing season. The mesh grid was judged to be suitable for wine grape production when the summer precipitation was in the optimum range from 255 mm to 1200 mm (Hannah et al., 2013). The winter temperature criterion was 15 of the monthly mean daily minimum temperature. The mesh grid was judged to be unsuitable when the winter temperature was below 15. The final mesh grid was judged to be suitable for wine grape production where all three criteria were satisfied. 3. Results and Discussion 3.1 Past to future shift of climatic suitability for Vitis vinifera The shift in climatic suitability for cultivating V. vinifera from the 20th to 21st century, as indicated by MIROC 3.2 Hires (section 2.1) data, is described in this section. Suitable summer temperature areas to cultivate V. vinifera in Hokkaido, i.e., where the temperature was more than 13.1, increased in size over time, as shown in Fig. 2b. Especially during present times, that is, the 1990s, 2000s, and 2010s, the expansion of the climatic suitability was significant. Before the 1990s, the suitable area size in Hokkaido had not changed much (data not shown in this manuscript) because summer temperatures displayed little change, as indicated by data from eight locations (Fig. 3). The area where summer precipitation was more than 1200 mm in Hokkaido was quite small during the 1990s and few changes are expected for these conditions throughout the 2070s (Fig. 2c). Although the area that failed to satisfy the winter temperature criterion was large in Hokkaido during the 1990s, that area dramatically shrank over the course of time (Fig. 2d). Fig. 3. 2b. Long-term variation of summer temperatures at the locations shown in Fig. 1. The background colors are the same as in Fig

4 Journal of Agricultural Meteorology 72 (3-4), 2016 In total, the suitable area for wine grape production, that is, where all three criteria were satisfied, expanded in size from the 1990s to the present (2010s), and the suitable area will cover almost all areas by the 2050s (Fig. 2a). While the suitable area in Hokkaido during the 1990s was limited in the southwestern area as the summer temperature range was under 14.4 (purple area in Fig. 2a), the area with temperatures higher than 14.4 (light blue area in Fig. 2a) expanded during the 2010s. This change corresponds well with the observed variety changes in Hokkaido, as mentioned in the Introduction (section 1). For instance, the Fig. 4. Differences among nine outputs for climatic suitability in Hokkaido that were obtained by using the future RECCA dataset. The nine outputs are given for combinations between three GCMs (MIROC, MPI, and NCAR) and three RAMs (JMA, RSM, and WRF). a) The mesh grid was judged to be suitable for wine grape production where all three criteria were satisfied. b) Summer temperature: white area shows the unsuitable area due to temperatures lower than c) Summer precipitation: black area shows the unsuitable area due to precipitation in excess of 1200 mm. d) Winter temperature: black area shows the unsuitable area due to temperatures lower than

5 M. Nemoto et al. Climatic suitability for wine grape production in Hokkaido lower temperature limits of Müller-Thurgau and Pinot Noir are 13.1 and 14.0, respectively (Hannah et al., 2013), and Pinot Noir can be stably cultivated in some areas of Hokkaido now. The climatic suitability for such varieties will cover almost all of the Hokkaido region by the 2050s and areas with summer temperatures higher than 17.0 will emerge. In accordance with the increasing summer temperatures, vineyard farmers might have to change to more heat-tolerant varieties. Figure 3 shows comparative information for eight places in Hokkaido (see Fig. 1 for the locations). The range of summer temperatures in these locations was approximately 2 throughout the 20th and 21st centuries. Southernmost Okushiri is the hottest among the eight places studied, followed by Mikasa and Yoichi, which are approximately the same, and then Furano. The other places (Rankoshi, Nayoro, Ikeda, and Kitami) have been cooler than the lower temperature limit for cultivating V. vinifera up to the 1990s, but these places are expected to get hotter after the 2000s. As Kanno (2013) indicated, a pronounced climate shift occurred in 1998, and since that time, global climate change has altered the meteorological environment for cultivating V. vinifera in Japan. All places are expected to stay within the optimal range throughout the 21st century. 3.2 Uncertainty of future predictions In this section, we focus our attention on Hokkaido where suitable area to cultivate V. vinifera is expected to expand significantly in the future, and we investigate the uncertainties of future climate projections by using the RECCA dataset (described in section 2.2). Figure 4 shows the variability of climate model results when global mean temperatures increase by 2 in comparison to the 1990s. The summer temperatures (Fig. 4b) were approximately the same in nine model runs. The summer precipitation (Fig. 4c) showed some variability in the southwestern part of Hokkaido, and there is a large possibility that it will increase compared to present conditions (Fig. 2c: 2010s). Regarding the winter temper- atures (Fig. 4d), the DDS results for the combination of NCAR and RSM model runs indicate that the areas lower than 15 C are distributed at high elevations, but the other models indicate that almost all land area in Hokkaido will satisfy the winter temperature criterion. Overall, unsuitable area remains in some model combinations, especially in the southwestern area of Hokkaido, which can be mainly attributed to high summer precipitation (Fig. 4a). Figure 5 summarizes the nine results for the future predictions described in Fig. 4a; these data show the number of future projections out of the nine RECCA projections that do not satisfy any of the three climatic criteria (summer temperature, summer precipitation, and winter temperature). The white area in Fig. 5 represents the area where the number of unsatisfied projections was zero, which means that all three requirements were satisfied in all nine models and the results for the suitability of cultivation of V. vinifera were associated with a high level of confidence. The yellow area in Fig. 5 represents the area where the number of unsatisfied projections was small (1 3), which means that the results for the suitability were associated with a small amount of uncertainty. The red area in Fig. 5 represents the area where the number of unsatisfied projections was large (4 9), which means that the results for the suitability were associated with a large amount of uncertainty or that the results for the unsuitability had a fairly high level of confidence. Therefore, the white areas in Fig. 5 can be thought of as robustly climatic suitable areas, given the changing climatic conditions and the uncertainties of the climate models, for the cultivation of V. vinifera. The yellow areas also can be thought of as highly climatic suitable areas. In the future, almost all low elevation areas in Hokkaido, except for areas near mountainous regions, were white and yellow areas in Fig. 5. These data suggest that future climate change conditions, in particular, the warming winter temperatures, will contribute positively to the cultivation of V. vinifera in Japan. Fig. 5. Stable areas for grape cultivation determined with consideration of the uncertainty of climate model predictions. The future represents the decade when global mean temperatures increase by 2 in comparison to the 1990s according to each GCM. The circles are the places for grape production selected in Fig

6 Journal of Agricultural Meteorology 72 (3-4), Conclusion Our results indicate that the present climate of Hokkaido is partly in a suitable range for the cultivation of V. vinifera, and the suitable area will expand in all low elevation areas in the future when global mean temperatures increase by 2 in comparison to the 1990s. Especially after the 1990s, expansion of climatic suitability for grape growth was significant. These findings were robust even given the uncertainty of climate projections. In future studies, it may be useful to explore changes in snow cover and frost to determine if these factors will impact the cultivation of grapes in Japan. Vineyard farmers in Hokkaido actually lower grape trees during the winter season to protect them from severe cold temperatures under the accumulated snow; therefore, it is possible that changes in snow cover (e.g., depth and duration) could affect growing conditions in Hokkaido, Japan. Moreover, the occurrence of a frost before or after the snow season can have destructive effects on grape trees. Probabilistic risk assessments that consider interannual variations in weather conditions including snow and frost, such as the one performed by Nemoto et al. (2012) for rice cropping, will be needed to improve the accuracy of future forecasts for grape cultivation in Japan. References Hannah, L., Roehrdanz, P. R., Ikegami, M., Shepard, A. V., Shaw, M. R., Tabor, G., Zhi, L., Marquet, P. A., and Hijmans, R. J., 2013: Climate change, wine, and conservation. Proceedings of the National Academy of Sciences of the United States of America, 110, Hasumi, H., and Emori, S. (eds.), 2004: K-1 coupled model (MIROC) description. K-1 Technical Report No.1, CCSR, University of Tokyo, 34 pp. Inatsu, M., Sato, T., Yamada, T. J., Kuno, R., Sugimoto, S., Farukh, M. A., Pokhrel, Y. N., and Kure, S., 2015: Multi-GCM by multi-ram experiments for dynamical downscaling on summertime climate change in Hokkaido. Atmospheric Science Letters, 16, Kanno, H., 2013: Strongly negative correlation between monthly mean temperatures in April and August since 1998 in northern Japan. Journal of the Meteorological Society of Japan, 91, Ministry of Agriculture, Forestry and Fisheries, 2015: Statistics of Agriculture, Forestry and Fisheries. < zyu/pdf/syukaku_ninasi_14.pdf> (accessed on 22 August 2015) (in Japanese). Nagata, O., Kobahashi, K., Niwa, K., Hirakawa, A., Takizawa, N., Ono, S., Yazaki, T., and Hirota, T., 2014: Report for the organized session of the annual meeting of the Society of Agricultural Meteorology of Japan 2014 Future perspective regarding the Sorachi area in Hokkaido as a place of wine production. Climate in Biosphere, 14, D18 D28 (in Japanese). Nakicenovic, N., and Swart, R., 2000: Special Report on Emissions Scenarios: A Special Report of Working Group III of the Intergovernmental Panel on Climate Change. Cambridge University Press, Cambridge, UK, 612 pp. Nemoto, M., Hamasaki, T., Sameshima, R., Kumagai, E., Ohno, H., Wakiyama, Y., Maruyama, A., Goto, S., and Ozawa, K., 2012: Probabilistic risk assessment of the rice cropping schedule for central Hokkaido, Japan. Journal of Applied Meteorology and Climatology, 51, Okada, M., Iizumi, T., Nishimori, M., and Yokozawa, M., 2009: Mesh climate data of Japan ver. 2 for climate change impact assessments under IPCC SRES A1B and A2. Journal of Agricultural Meteorology, 65, Teutschbein, C., and Seibert, J., 2012: Bias correction of regional climate model simulations for hydrological climate-change impact studies: Review and evaluation of different methods. Journal of Hydrology, 456,