论文摘要 |
Urban heat island effects are created by large cities, and weather data collected in or near heat spots may be inappropriate for crop growth simulation. Research was conducted to evaluate the effects of different sets of historical weather data on DSSAT (Decision Support System for Agrotechnology Transfer) and APSIM (Agricultural Production Systems sIMulator) model simulations for winter wheat (Triticum aestivum L.) production in the North China Plain (NCP). Yield data from 10 recent years and three locations in the NCP were used for model calibration and validation. Three weather datasets including data from the exact experimental site, a nearby town and a nearby large city were used to obtain three sets of model parameters for each location. The different model parameters were further used to assess crop performance on regional scales. The well-calibrated and validated APSIM and DSSAT gave nearly identical average regional yields, but the model parameters in DSSAT were more sensitive to weather sources. Using the parameters derived from the source of city weather data tended to overestimate the yield in higher latitudes and underestimate the yield in lower latitude regions due to the higher thermal units and lower growth rates in the model parameters, which prolonged the growth duration. The results from this study also indicated that to reduce the bias in regional crop yield simulation, increasing the sites for model calibration should be recommended. During model simulation, the changes in model parameters to suit the source of the weather data increased the stability of the model performance but decreased the model sensitivity in responding to the changes in growing conditions. |