Objective:Tuberculosis (TB) is still one of the most serious infectious diseases in mainland China,ranking the second place only behind the virus hepatitis.Although several studies have explored the TB incidence and its associated factors,they roughly analyzed it at the provincial level and scarcely paid close attention to meteorological factors.This study aimed to detect the spatial and temporal distribution patterns of TB incidence in mainland China during 2005-2015 and to assess the association between TB incidence and meteorological factors at a prefecture level.Methods:An ecological study was conducted using data on patients diagnosed with TB in mainland China,for the period 2005 to 2015.Descriptive epidemiology methods were conducted to summarize the epidemiological characteristics of TB firstly.Then Morans I index and Getis-Ord Gi index were used to detect the spatial autocorrelation and hot spots in 340 prefectures from 2005 to 2015.Finally,the relationship between TB incidence and meteorological factors was explored by geographically weighted regression (GWR)model at a prefecture level.Results:Temporal characteristic analysis showed obvious seasonal variation laws of TB cases with a peak in March or April and a trough in January or February during 2005-2015.In addition,the total number of TB cases and TB incidence declined in general from 2005 to 2015.The spatial distribution for TB incidence presented that it was relatively low in the eastern parts and high in the western parts of mainland China.Spatial autocorrelation analysis (Z=18.142,P<0.001) and hotspot analysis of TB incidence found the similar results that the western China especially Xinjiang Uygur Autonomous Region and Tibet Autonomous Region had the high-high cluster and hotspots regions,which should be the key areas of health resource allocation.Local coefficients of average rainfall (AR) from 2005 to 2015 had a positive correlation with TB incidence while average relative humidity (ARH) and average sunshine duration (ASD) had a negative correlation with TB incidence in all of the 340 prefectures ranging from 4.27 to 26.45,-19.76 to -10.08 and -31.71 to -2.55,respectively.However,the relationships between other three meteorological factors and TB incidence in 340 prefectures were not completely consistent in mainland China.Average temperature (AT),average air pressure (AAP ) and average wind speed (AWS) were all negatively associated in most of the areas except some small regions in east China with the local coefficients ranging from -28.71 to 12.84,-7.78 to 10.49 and -8.58 to 2.99,respectively.The adjusted R2 of GWR model was 0.3523,which meant the model could explain 35.23% of the total variance in TB incidence.Compared with the Ordinary Least Square Regression (OLS) model (adj R2 = 0.1623,AICc= 3227.48),GWR model improved the fitness of regression (adj R2 = 0.3523,AICc=3146.56).Conclusion:In summary,the TB incidence in mainland China presented significantly temporal variability and spatial variability.Regional TB monitoring planning and prevention and control strategies should be formulated based on the spatial characteristics of TB incidence and spatial variability of the local relationship between TB incidence and meteorological factors.