Characterizing Urbanization over two decades in the Pinga Oya River Catchment in Central Sri Lanka using Land Surveys and Google Earth images for Application for Dengue and Hydrological Risk Assessment

Abstract:

Asia is urbanizing rapidly which is leading to a rise in flooding risk, low flows and the incidence of dengue. In Southern-most South Asia, peri-urban Sri Lanka is facing increasing challenges from disasters, infectious diseases and air pollution. We are undertaking flood risk and dengue risk assessment which requires village level characteristics to be able to model risk . Given the rapid urbanization, it requires quantification of the constructed areas, bridges, river channels and land use density. Spatial information at high resolution is available from land surveys conducted at 1:1000 from the Sri Lanka Department of Survey. Time series of free remotely sensed images on land cover is available through Google Earth and similar portals.

We are working on flood risk assessment and dengue risk assessment for a small catchment - Pinga Oya - off the main Mahaweli River in a hilly location. Flooding has been increasingly frequent even with lower rainfall thresholds starting this century. Dengue incidence has peaked spatially on occasion in this catchment. Here, we assess the possibility of time series of Google Earth satellite images (from 2003 to 2019) along with 1:1000 digital surveys done by the Department of Survey to assess changes in land use and the built environment around Pinga Oya.

The work involved aligning the digital surveys in Auto-Cad software with the pixel data from Google Earth and the implementation of feature detection algorithm to obtain the geometrics of the structural features. We were able to quantify (a) the changing number of buildings including extensions, (b) changes in the number of structures in the river and close to it - including bridges and retaining walls and (c) the extent of urbanization by user defined selected area.

We are incorporating these results into the analysis of dengue transmission along with entomological, climatic and demographical information. We are also assessing the impact of land use change on rising incidence of floods and low flows in the river catchments.

The methodology of combining freely available precise ground based surveys along with time series of remotely sensed images is demonstrated to be useful in characterizing urbanization and features at fine scales over the last two decades.