中文摘要：

目的探讨乙肝发病的时空分布特征并预测其发病趋势,分析原因及其影响因素.方法运用GIS和数学模型对发病率数据进行分析,将空间聚集现象及地理可视化相结合,并建立ARIMA预测模型.结果南昌市乙肝发病率在2006-2012年总体呈现下降趋势,无明显季节性变化;南昌市乙肝发病率呈聚集性分布,高值聚集区主要集中在东部、低值聚集区集中在中部偏西北地区,发病率空间分布格局不稳定,在“西北-东南”方向发病率逐级递增,同时存在“热点区域”从东北向东南方向转移的趋势;ARIMA（1,0,1）（0,1,1）12模型拟合效果最好,残差通过白噪声检验.结论乙肝发病在时间上聚集不明显,在空间上呈聚集性分布;ARIMA预测模型拟合精度高,能够为乙肝的早期监控预警提供建议.

英文摘要：

Objective ： To explore temporal and spatial distribution characteristics of hepatitis B, predict incidence trend, and analyze the results and influencing factors. Methods：Using geographical information technology and mathematical modeling, combining spatial ag-gregation phenomenon and geo visualization, and establishing ARIMA prediction model. Results ; The hepatitis B incidence of Nanchang city between 2006 and 2012 presents the overall drop without obvious seasonal variation. The hepatitis B incidence of Nanchang city presents aggregation distribution. High value aggregation areas mainly focus on east region, while low value aggregation areas inversely focus on middle region to the northwest. Spatial distribution pattern of hepatitis B incidence is unstable, the hepatitis B incidence in-creases on wnorthwest - southeastM direction by degrees, and the results show the trend of metastasis of hot area from northeast to southeast. And the best model is ARIMA（ 1,0,1）（0,1,1）12, and white noise was used to test residual. Conclusion ： Temporal distri-bution characteristics hepatitis B incidence are unobvious, while it＇s aggregated in spatial distribution. ARIMA prediction model has high fitting accuracy, and be able to provide evidence for monitoring and early - warning of hepatitis B.