A review on statistical models for identifying climate contributions to crop yields

摘要: Statistical models using historical data on crop yields and weather to calibrate relatively simple regression equations have been widely extensively applied in previous studies, provided a common alternative process-based models, which require extensive input cultivar, management, soil conditions. However, very few studies had conducted review systematically the statistical for indentifying climate contributions yields. This paper introduces three main methods, i.e., time-series model, cross-section model panel used identify such issues field of agrometeorology. Generally, research spatial scale could be categorized into two types including site regional (e.g. global scale, national provincial county scale). Four exist identifying response sensitivity change by models. The include extent temporal non-climatic trend removal, colinearity existing variables non-consideration adaptations. Respective resolutions above four put forward section perspective future finally.