Abstract:     A process-oriented methodology to conduct precise evaluation temporally and spatially on temperature suitability for potato growth was applied in China. Arable lands in China were gridded with 1 km×1 km geographic units, and potential potato phenology in each unit was automatically identified in terms of the potato planting initial temperature and effective accumulated temperature. A temperature thermal response coefficient model was used to compute a temperature suitability value for each day of potato phenology in each geographic unit. In addition, five temperature suitability ranking methods were applied to define suitable areas: (1) upper fourth quantile, (2) median, (3) expected value+1/4 standard deviation, (4) expected value+1/2 standard deviation, (5) expected value+1 standard deviation. A validation indicator was innovated to test the effectiveness of the five ranking methods. The results showed that from a strict degree point of view, the five methods sequence was as follows: 1=3>4>2>5, with a and c determined as the two best ranking methods. For methods 1 and 3, the suitable potato growing area was 1 of 57.76×10⁴ km². In addition, the suitable areas were spatially coincident with the main potato producing counties. The study output technically supports the proposal from China’s government that there is a large potential area to grow winter-ploughed potato in South China because the potential suitable area for growing potato is approximately 2×10⁷ ha. In southeast Heilongjiang and east Jilin, where it is hilly and mountainous, there are still some potentially suitable areas for potato growing accounting for nearly 2.32×10⁶ ha. The authors suggest to optimize the agricultural regionalization and layout in China and to adjust the cropping pattern structure.

Key words: potato ,  temperature suitability ,  temperature thermal response coefficient model ,  ranking method ,  validation