A Web GIS Application for decision support in airstrip allocation and serviceability impact on rural areas in Papua New Guinea. Follow link to access webapp https://niuginitravelor.shinyapps.io/Airport-Inspector/
There are about 9.3 million people as of June18 2022 within Papua New Guinea with bulk of the population residing in remote parts of the country. Many of these places lack basic government services due to several compounding factors such as: rugged terrain, remoteness, limited economic activities, government funding, development focus areas, outdated demographics and inefficient service intervention sites. In addition to that, Papua New Guinea lacks research publications or reports that are accessible to make sound decisions for site intervention measures. Airstrips alone cover much of the country however, there is no way of measuring just how much of the country does it serve these rural populations. Questions like how far does a village have to be in order for it to benefit from proximity to an airstrip? Or, what areas of the country are segregated from airstrip serviceability? Responses to these questions are varied and limited to local area knowledge hence, introducing uncertainties which will have an impact on decision making. Therefore, there is a need for a platform that can quantitatively measure and visualise airstrip serviceability areas for the country and function as a support tool for decision making. The development of the Airport Inspector App which is a Web GIS implementation endeavours to quantify how much of an impact an airstrip provides in terms of serviceability on a provincial or national scale. This app is perhaps the first of its kind that creates an automation workflow with real time changes to modelled outputs whilst keeping it simple for ease of use by non-gis users. Most of the GIS workflow have been coded with computer programming language R to perform the automation while only requiring simple user input such as uploading spatial files or input numerical variables. It is simple intuitive and has the potential to be a powerful tool to target areas or discover areas that lack serviceability hence, adding value to funding proposals that incorporate the Airport Inspector outputs.
This tool functions as an image processing tool that displays panoramic views of the airstrip. These are 360 degree image files which are geotagged to the location it was taken. The tool provides an immersive environment that has the same aesthetic look and function of google streetview. Currently under development to include an upload button to upload these 360 degree image files.
The Location Prediction Tool works by taking input coordinates in decimal degree (dd) format along with elevation in feet (ft) to predict the most likely province that it is located in Papua New Guinea. The tool was developed using a machine learning approach in which the Randon Forest algorithm was used. It was trained on 475 data points of airstrip locations spread around the country attaining a 90% precision in its prediction. An additional feature of the LPT is the ability for it to obtain population statistics of the province through web scarping, which is a technique for obtaining data stored on the web. Note that any coordinate input that do not fall within Papua New Guinea will be inaccurate in its prediction therefore caution must be taken when choosing coordinates.
The Predictive Aerodrome Reach and Impact Area Model creates an interpolated surface from gps location of airstrips stored in its database. This surface will be weighted to topography, settlement locations and waterway connectivity to provide a quantifiable impact and reach assessment of an area if an airstrip is operational or allocated anywhere in Papua New Guinea. For this tool to perform its analysis, it requires a country spatial file of Papua New Guinea and the province to perform its analysis. The output is the modelled reach of airstrips within the region where it assumes the proximity distance people prefer to travel.
- The Airport Inspector for the first time, makes it possible to visualize serviceability into these remote areas in the country and provides a glimpse of just how effective airstrip location can make to rural areas.
- Provides a simple interface for any user with limited GIS experience.
- Map Outputs and statistical results can be incorporated into reports, adding value to data driven proposals for funding, awareness etc
- Can be used to allocate areas for proposed intervention measures e.g, the PARIAM identified an area that is out of reach from nearby airstrips. Therefore an intervention measure would be to build an airstrip there.
- Can help the organisation measure its success through the PARIAM tool by showing how much of the country it is impacting through its airstrip programs.
- Feature Detection tool is a concept tool to detect cones on the airstrip and return the total number of cones as well as the colours detected from a flyby image either taken from a drone or from the window of an aircraft or helicopter. This would improve survey time during maintenance or survey runs.
- Soil Testing Tool is a concept tool that endeavours to detect and predict areas of the airstrip that are soft, firm or solid. This will be built using a supervised machine learning approach that will take variables such as vegetation index, soil index and penetrometer results to create the predicted surface for every part of the airstrip. It hopes to use drone imagery that’s captured using a multispectral camera.