Remote Sensing Diagnosis Of Dengue Fever In Nepal

Dengue fever is a mosquito-borne viral disease which is transmitted from one person to another through bites of female Aedes spp.mosquito.In the recent years,it has been a major public health problem of the tropical and subtropical countries throughout the world.While the incidences of dengue fever has risen sharply on one hand,the disease has expanded in new areas on the other.In Nepal,dengue fever was first reported in 2004 AD.Since then,the disease has spread rapidly covering wide geographical areas and frequent outbreaks occurred in different parts of the country.Occurrence and spread of dengue fever depends on several environmental and socioeconomic factors resulting various spatiotemporal patterns of the disease.However,little is known about the spatial and temporal characteristics of the disease in Nepal.The main aim of this research was to improve understanding on the spatiotemporal dynamics and associated risk factors and forecast future disease risk under the climate change scenarios.This dissertation has utilized multi sourced data and multiple analyses to achieve the research objectives.Dengue data were primarily collected from the Epidemiology and Disease Control Division(EDCD)of the Government of Nepal.Earth observation data were collected from Landsat,Moderate Resolution Imaging Spectroradiometer(MODIS)and Tropical Rainfall Measuring Mission(TRMM)satellites.The dissertation has also utilized worldclim bioclimatic layers,Shuttle Radar Topographic Mission(SRTM)Digital Elevation Model(DEM)and WorldPop population data.Socioeconomic data were collected from the Central Bureau of Statistics(CBS)of Government of Nepal.Chorpleth mapping technique and SaTScan method were employed to map the distribution and to asses statistically significant clusters of the disease.Semi-parametric geographically weighted regression model was applied to assess spatial variation of dengue fever incidence and its associations with potential environmental and socioeconomic risk factors.Temporal associations of dengue fever and remotely sensed environmental variables were assessed based on Poisson generalized additive model and the cross-correlation analysis.Maxent ecological niche model was used to map the climatically suitable areas of dengue for the present and for the future in the context of climate change.The result revealed the presence of substantial district level spatiotemporal variations of dengue fever in Nepal.Higher incidences were observed in the southern lowland districts compared to other parts of the country.Primary and first secondary clusters were detected in Chitawan and Jhapa districts respectively.Temporal analysis showed significant inter annual and seasonal variations of the disease which follow the pattern of monsoon rainfall.Ward level(local scale)analyses revealed highly heterogeneous distribution of dengue fever in Jhapa district,Nepal.Statistically significant clusters were discovered in the highly populated urban centers.Semi-parametric geographically weighted regression(s-GWR)model identified proportion of urban area,proximity to road and population density as the major risk factors for the spatial variation of dengue in local scale.The s-GWR model best explained spatial variation followed by GWR and OLS model with minimum AIC and Maximum deviance explained.Temporal associations of dengue fever with satellite derived environmental variables revealed that dengue fever is significantly associated with satellite estimated precipitation,normalized difference vegetation index(NDVI)and enhanced vegetation index(EVI)synchronously and with different lag periods.However,the associations were weak and insignificant with daytime land surface temperature(dLST)and nighttime land surface temperature(nLST)immediately but significant after 4/5 months.Precipitation,d LST and NDVI together best explained the temporal patterns of dengue fever cases.The best fit model significantly improved when the model was adjusted with delayed effects of the selected variables.The delayed effects adjusted best fit model predicted dengue cases with a high degree of accuracy which was concurrent with observed and 10-fold cross validation.The cross correlationThis study has also mapped climatically suitable area of dengue fever and assessed the impact of climate change on it based on Maxent Ecological Modelling.Simulation based estimates suggest that climatically suitable areas of dengue fever are presently distributed throughout the lowland Tarai from east to the west and less elevated river valleys.Under the climate change,these areas will be slightly shifted towards higher elevation.As a result,proportion of population exposed with dengue fever will be increased in the future.While this study was able to enhance knowledge on spatiotemporal dynamics of dengue fever based on the available data and time constraints,the approach could be further extended with higher resolution spatial and temporal data.The nonlinear effects of environmental and metrological factors on the temporal pattern of the disease and effects of climate change in temporal risk window is also recommended for the future research.In conclusion,the dissertation demonstrates a number of novel contributions to enhance understanding in spatial and temporal epidemiology of dengue fever in Nepal.More specific innovation of this dissertation are as follow.1)The distribution of dengue fever exhibits distinct spatiotemporal variation in the district and ward level in Nepal.2)Such spatial variations is influenced by a range of environmental,socioeconomic and proximate variables.Semi parametric geographically weighted regression(s-GWR)model better explain the transmission dynamics of dengue fever and its spatial risk factors compared to the global Ordinary Least Square(OLS)and Geographically Weighted Regression(GWR)models.3)This dissertation showed the importance of remotely sensed environmental variables to explain temporal patterns of dengue fever.The delayed effects identified in this research can be used to develop remote sensing based early warning system.4)The application of ENM in understanding the geographical and landscape epidemiology of dengue fever was also demonstrated.This dissertation defined for the first time the present and future spatial of climatic niche of dengue fever in Nepal.These findings may assist in improving the guidelines for setting priority areas for Aedes surveillance and effective allocation of scarce resources.