Disaster Impact Assessment Using GIS and Remote Sensing
Disaster management is all about knowing the right data, at the right time and initiating correct preventive and handling measures in moments of pressure. Optimum understanding of area, right amount of pre-planning and effective mobilization of during and post event resources all contribute towards mitigating effects of calamity. With the rising population, ever increasing dense urban areas, extreme climate conditions make the entire system more and ever gravely vulnerable to disaster events. Classification of natural calamities as disasters in today’s day and age would not cover the entire spectrum. Events like large-scale leak of bio-hazards, epidemics, failure of civic amenities could also make the cut. Having a deep understanding of critical areas along with strengths, weaknesses, vulnerabilities and location insights of the area of interest can help authorities better plan, prepare and execute during any event of disaster. In this article we will focus on naturally occurring disasters and how remote sensing and location intelligence play a key role in the entire life-cycle of a disaster management plan.
- Prepare, respond, recover, mitigate are the four major steps in the entire disaster management scenario and geo-spatial data input plays a vital role at all stages in every step.
- Prepare: Model the area in 3D, model vulnerability, predict risk zones based on historic climatic conditions, prepare for worst situation, and plan most effective routes, relief locations
- Respond: Live track the situation with integrated sensor input, actively work on extraction plan and track effectiveness of Standard operating protocols
- Recover: Assess expanse of damage, create impact assessment reports, and respond to incoming field level damage reports, validated field reports
- Mitigate: Mitigation phase more tied with the effectiveness of preparation and response phase. Effective mitigation leads to hazards not turning into disasters and also contributes to effective planning and future preparedness.
Vasundharaa has closely worked in the impact assessment phase thus far and has recently been tasked by a couple of affected ULBs to present a project report on future preparation models, predictive models and data analysis.
Forest Fires:
Forest fires, a phenomenon that is very much a man-made problem especially in India. Hundreds of Sq.km for natural flora and fauna get burned, and it looks like an inferno. There have been major fire events this year around the world, with the largest reported in the Amazonian forests. A similar large scale fire was reported in the Bandipur national park in Feb 2019. The satellite images in the shortwave infrared bands, not only aid in monitoring the event on a near real-time basis as seen in Figure 1 (Left and Right), where the bright orange indicates an active fire but also post event assessment as seen in Figure 2. The area of the burnt patch was extracted from the images to be around 83.09 km2. The vector of the burnt area was extracted automatically using our algorithms and can be delivered to the authorities in a very small time. This methodology has been proposed to the Forest Survey of India, with the aim to enhancing their fire alert system.
Floods:
Geo-spatial data plays a vital role in all the aspects of disaster management during a flood event. Represented in the video is an animation of predicted submergence for the city of Sangli. This forms the base information of all types of risk zone planning, emergency route planning, relief area planning and mitigation during the event. Real scale 3D models can be generated and flow simulations can help understand which area has highest risk are.
Figure 3: SAR Images of Bihar 2019 (Left: Pre-event Image) (Right: Post-event Image), Source: Sentinel Hub
Another challenge for the data analyst especially during the floods seasons is the prevailing clouds. Clouds make earth observation using optical sensors very difficult, and need is to observe using active Synthetic Aperture Radar (SAR) sensors. This class of very specific sensors can penetrate clouds and have a very strong signature with water. Signatures of changes in surface water can be effectively observed and forms a very important dataset for impact assessment. Shown here are images for Bihar floods. Images in Figure 3 show the pre (Left Image) and post (Right Image) flood situation near Patna and Begusarai, Bihar. Figure 4 shows the comparative submergence patterns, with red areas being the areas with full submergence. The GIF (Figure 5) shows comparison between pre and post floods.
Other aspects of importance of geo-spatial resources is mapping and tracking rainfall, local climate conditions, monitoring surface soil moisture for predicting fire vulnerable areas, integrate citizen reports and social media postings for rapid and live situation updates. Remote sensing thus is very important in studying, monitoring and predicting natural disasters. In a nutshell, remote sensing with geospatial data offers high potential in dealing with an on-going event and provide vital information on the spread, magnitude and direction of the event, serve as a support system to plan and execute rescue operations in a quick and timely manner.
