Exploration of method for discrimination between trend crop yield and climatic fluctuant yield

Abstract

The impact of climate change on crops yields had become one of the hotspots in the research on global climate change.The empirical researches of long time-sequence analysis were effective approaches for analysis of relations between crops yields and meteorological factors.In most of empirical researches,the long time-sequence crops yield is generally decomposed into trend yield(which fits the real crop yield by a trend line depended on mathematic model and which is considered as a result of agricultural technology development and agriculture investment),climatic fluctuant yields(which are considered as contribution of climate fluctuation) and random error.As different trend yields come from different mathematic fitting model,it would get different climatic fluctuant yields,even diametrical result(3-point moving average and 5-point moving average may have negative effect in some years).It is the critical steps that how to simulate the trend yield and separate the trend yield to get an correct or more accurate climatic fluctuant yield in such studies.In this paper,a series rules were put forward to show how to choose a more accurate fitting model to calculate trend yield,which may overcome the disadvantage of random choosing models.Take climate change effect on cotton output for an example,the paper wants to explain how to choose fitting model to get an correct trend yield.Puts forward three rules for method selection:trends yield simulation curve should accord with the process of social technology development reality,the variation characteristics of trends yield should have the same or similar trend in whole research region or country(namely the development of social and technical level no significant difference),different administrative regions which have similar climatic conditions should have strong correlation,and the key climate factors of model obtained which affect the research object(crop) growth should be consistent with the crop reproductive characteristics and crop response to the climate factors.