resilience to natural disasters: advances in earth ... · the sendai framework made specific...
TRANSCRIPT
Resilience to natural disasters: advances in Earth Observation;
knowing more to lose less
Seventh Arab-American Frontiers symposium
Cairo, Egypt, November 17-19, 2019
George Mitri, Ph.D.
Institute of the Environment,
University of Balamand
Structure
Background
Case for space
Seizing the opportunity
A case from Lebanon
Take-home messages
Background
Source: © 2019 Munich Re, Geo Risks Research, NatCatSERVICE. As of March 2019.
Three of the UN Sustainable Development Goals include disaster resilience as a target (i.e., ‘no poverty’,‘sustainable cities and communities’, and ‘climate action’).
The UN has established the Sendai Framework for Disaster Risk Reduction (i.e., leading to the ‘substantialreduction of disaster risk and losses in lives, livelihoods and health and in the economic, physical, social,cultural and environmental assets of persons, businesses, communities and countries’)
The Sendai framework made specific recommendations about the use of Earth Observationsolutions to address disaster resilience, including:
‘to promote real time access to reliable data, make use of space and in situ information’
‘to promote and enhance…communications and geospatial and space-based technologiesand related services; maintain and strengthen in situ and remotely-sensed earth and climateobservations’
Developing countries lack the infrastructure to gather and process data needed to predict when and where natural disasters might occur and what the impact would be
Capacity to respond to the disaster is often overwhelmed. Terrestrial communication networks can often be damaged during disasters
Limited information, visibility and communications around post-disaster population movements and infrastructure damage (especially in developing countries).
EO responds to the need of impactful and cost-effective approaches to prepare, respondand recover from disasters (especially in developing countries).
EO supports a robust insurance market through improved calculation of risk:
‘Over 40% of economic losses are insured in developed countries, compared to 10% and5% in middle and low-income countries respectively’.
EO data contributes to precisely identifying and estimating the value of assets (i.e., homes,infrastructure) to price and offer insurance products.
The case for space
Space solutions support disaster resilience
Number of annual publications on remote sensing-based proxies for
Disaster Risk Management (DRM) (source: Ghaffarian et al., 2018)
Geostationary (GEO): satellites are located 36,000 km above the Earth. The fixed positionsof satellites in GEO provide for regional telecommunications services including: voice,video and broadband data; and weather services.
Low Earth Orbit (LEO): below 2,000 km above the earth. Satellites continually orbit theearth to allow global coverage within a ‘revisit period’ providing a variety of earthobservation services.
The detail discernible in an image is dependent on the spatial resolution of the
sensor and refers to the size of the smallest possible feature that can be
detected.
Temporal resolution varies by satellite and describes the time it takes for an
individual satellite to orbit and revisit a specific area. Some satellite operate as a
constellation with multiple satellites working together to increase their global
coverage
Charters and emergency services for use of space solutions in disaster resilience
They establish the necessary connections between data providers, information developers,and end users to ensure that decision-makers can benefit from satellite technology in thedisaster resilience community
The InternationalCharter for Spaceand Major Disasters:This charter is aglobal mechanism forcountries to accesssatellite imagery insupport of theirdisaster responseactivities. The chartercan provide rapidaccess to free datafrom a virtualconstellation ofsatellites owned byspace agencies andsatellite operators.
Copernicus Emergency Management Service (Copernicus EMS): Provides rapid mapping andinformation for emergency response in relation to different types of disasters. These includemeteorological hazards, geophysical hazards, deliberate and accidental man-made disastersand other humanitarian disasters as well as prevention, preparedness, response and recoveryactivities.
Free space derived services available to all via the internet
Seizing the opportunity
Space solutions support disaster resilience
EO satellite data support each phase of DRR
Use of EO for Mitigation and Preparedness (examples)
Assessment of wildfire fuel loads enables effective hazardreduction burns to be conducted in high-risk areas.
Accurate information on topography is a key input to modelingrisks associated with inundation from floods and tsunamis.
Many of the factors governing the occurrence of drought can bemeasured using satellites and this supports the assessment ofvulnerability that enables mitigation and preparation.
Monitoring risk – weather hazards
Over 90% of disastersare linked tohydrometeorologicalhazards.
Satellites provideuniquemeteorological andenvironmentalobservations thatenable warnings ofextreme weatherevents on a globalscale.
Disaster response and recovery (examples)
Typhoon Haiyan in 2013: >1,000 high-resolution images provided (by the Charter) toassess infrastructure damage (e.g., structures, houses, oil facilities).
Satellite radar imagery is usually employed to produce flood-extent maps (Radarimagery enables all-weather, daynight response)
Images (often high resolution) provide detailed assessment of structural loss, damageto houses, and damage to critical transport infrastructure (e.g., roads, rail, andairports).
Long-term climate monitoring
Climate change amplifies risk factors
Monitoring long-term trends in climate variability isimproves our understanding of how this intensification willimpact weather-related DRM.
Use of EO for improved fire risk management: a case study from
Lebanon
Wildfires occur every year over large areas across different Arab countries in the Mena region: Morocco, Algeria, Tunisia, Lebanon and Syria
NASA's Aqua satellite detected heat signatures from fires and saw plumes of smoke from fires burning
in northern Algeria and Tunisia, along the coast of the Mediterranean Sea – August 4, 2017
A case study from
Lebanon showing how
remotely sensed data and
GIS analysis may be used
in all phases of a fire
management programme.
Communication and dissemination of fire riskMitri et al., 2014a
Fire danger forecast
Mitri et al. 2017Mitri, 2015Sakr et al 2010
Fuel type mappingMitri et al 2012Mitri et al 2011aGitas et al 2006aGitas et 2006b
Burned area mappingKatagis et al 2014aKatagis et al 2014bMitri and Gitas 2004Gitas et al 2004aMitri and Gitas 2004Mitri and Gitas 2003Gitas et al 2003Mitri and Gitas 2002
Towards integrated fire risk management
Fire type and severity mappingMitri and Gitas 2006Gitas et al 2006Mitri and Gitas 2005Gitas et al 2009Mitri and Gitas 2008
Vegetation recoveryEl Halabi, Mitri et al., 2014aEl Halabi, Mitri et al., 2014bFiorucci, Mitri et al 2013Mitri and Gitas 2013Mitri and Fiorucci 2012Gitas, Mitri et al 2012Mitri and Gitas 2010aMitri and Gitas 2010bMitri and Gitas 2007
Socio-economic mappingMitri et al 2016
Wildfire potential in present and future
Mitri et al, 2015a
Fire risk mapping/monitoring Mitri et al 2015bMitri et al 2014cSalloum and Mitri 2014Salloum Mitri et al 2013Mitri et el 2012Mitri et al 2011bSakr et al 2011Gitas et al 2004b
Pre-fire assessment During fire After fire
Fire detectionSakr et al, 2014
Fire hazard map
Topographic data
Vegetation structure
Fuel combustibility
Fire Potential Index (1x1 km of spatial resolution)
Spatial climatic data (current conditions)
Settlements vulnerability
Urbanized land
Environmental vulnerability
Protected areas, protected valleys, Natural reserves
Overall land vulnerability
Environment based fire risk
Socio-economic based fire risk
Social factors (e.g. density of population)
Overall current fire risk (results at the municipality level)
Economic factors (e.g. agriculture production)
Socio-economic changes factors (e.g. changes from forests to settlements)
Statistical analysis
Advancing fire danger forecast in Lebanon (simplified flowchart)*
Burned areas/fire occurrence
Environmental/geophysical aspect
Socio-economic aspect
Weather forecast
EFFISFWI
European Centre for Medium-Range Weather Forecast (ECMWF)
Mitri, G., Saba, S., Nader, M., and McWethy, D. 2017. Developing Lebanon’s fire danger forecast. International Journal of Disaster Risk Reduction. Vol. 24. Pp 332-339.
*Work supported by USAID-PEER project (2012-2016)
Hazard
Vulnerability
Risk
Adopted mapping techniques: the concept behind
Use of high resolution images > Employment
of advanced techniques in image analysis
close to human perception
Human perception does not simply register
Constructivism has a strong impact on the
field of artificial intelligence
Analytical approach adopted for processing satellite data
As the brain does, Geographic Object-Based Image Analysis (GEOBIA)
processes image information in an object oriented way.
The result of image analysis with GEOBIA comes in the form of a hierarchical
network of image objects (i.e., corresponds to the approach of constructivism).
Besides its neighbors, each object also knows its sub-objects and super-objects
in such a strict hierarchical structure.
The basic idea is to replace the
two strictly logical statements
“yes” and “no” (i.e., crisp system)
by the continuous range of [0...1],
where 0 means “exactly no” and
1 means “exactly yes.”
Fuzzy classification systems
are well suited to handling most
vagueness in remote sensing
information extraction.
Fueltype
% Coverage Description
1 Ground fuels (cover > 50%) Grass
2Surface fuels (shrub cover > 60%; treecover < 50%)
Grassland, shrubland (smaller than 0.3-0.6 m and with ahigh percentage of grassland), and clear-cuts, whereslash was not removed.
3Medium-height shrubs (shrub cover >60%; tree cover < 50%)
Shrubs between 0.6 and 2.0 m
4Tall shrubs (shrub cover > 60%; treecover <50%)
High shrubs (between 2.0 and 4.0 m) and young treesresulting from natural regeneration or forestation.
5
Tree stands (>4 m) with a cleanground surface (shrub cover < 30%)
The ground fuel was removed either by prescribedburning or by mechanical means. This situation may alsooccur in closed canopies in which the lack of sunlightinhibits the growth of surface vegetation.
6Tree stands (>4m) with mediumsurface fuels (shrub cover > 30%)
The base of the canopies is well above the surface fuellayer (>0.5 m). The fuel consists essentially of smallshrubs, grass, litter, and duff.
7Tree stands (> 4m) with heavy surfacefuels (shrub cover >30%)
Stands with a very dense surface fuel layer and with avery small vertical gap to the canopy base (<0.5 m).
Fire spread map of Lebanon
Spatial distribution of KBDI classes throughout the year (excluding
agricultural and un-vegetated areas)
Distribution of the average monthly KBDI in function of average mean
elevation (excluding agricultural and un-vegetated areas)
WUI area (ha) in risk of fire spread
WUI mapping results
Wildland-Urban Interface (WUI)
The WUI is described as the line, area, or zone where structures and
other human developments meet or intermingle with undeveloped
wildland or vegetative fuels (the United States Department of the
Interior, 1995)
WAI mapping results
The agricultural interface can be defined as an interface where farms, crops,
and orchard, irrigated or non-irrigated, are exposed to forest fires.
Focus on mapping demographic vulnerability (1)
The vulnerability of a territory is defined as the extent of loss or damage that may affectthe population, goods and environment, after a forest fire.
This vulnerability shall be determined through the integration of the Occupation,Boundary and Scatter indicators.
Focus on mapping demographic vulnerability (5): Overall demographic vulnerability
Demographic vulnerability = I.Occupation + 1,25 x I.Boundary + 1,25 x I.Scatter
Boundary and Occupation weighted by a factor of 1.25.
From the viewpoint of the demographic vulnerability to forest fires, the high scatter or boundary is more dangerous than the high occupation
Mitri, G., Saba, S., Nader, M., and McWethy, D. 2017. Developing Lebanon’s fire danger forecast. International Journal of Disaster Risk Reduction. Vol. 24. Pp 332-339.
“At the rate we’ve seen so far, this
year unfortunately will be very bad
in terms of the number of fires and
the areas where we see them,”
George Mitri, the director of the
Land and Natural Resources
Program at Balamand University,
told The Daily Star.
“It’s beyond the abilities of the state to
deal with it - the equipment and
resources aren’t enough - and when
fires in dense forests are coupled with
dry air and wind, the fire fronts can
get up to 20 meters high. It’s hard for
anyone to cope,”
Mitri added: “Prevention comes
before everything else, because if
these tree don’t catch fire today, they
will tomorrow.”
12 October 2019
15 October 2019: large number of fire occurrence across the country
EUCPM activation on October 15
The European Commission's Copernicus satellite mapping service was activated on October15 and two satellite maps were produced.
The alerts of European Forest Fire Information System on October 15 “the fire risk willremain high to extreme across most of the country.
By 17 October 2019: largenumber of fire occurrence (morethan 120 fires) across thecountry
More than 90% of fires in wildland occurred in area of moderate to very high fire risk
The most catastrophic fires burned in area classified as high to very high risk of fire
Three main take-home messages
Future challenges in respect to EO data analytics:
Fast growth of data from EO programmes
Big amount of raw data to be processed
Data need to be made available timely for improved disaster resilience
Increasing rely on Web-based workflows:
Analysts and end-users rely more and more on web-based workflows
General direction should allow users to directly process EO data web-based
Hidden treasure of data:
EO data captured regularly without knowing of what it might contain or used for especially in the
context of predicting and monitoring future disasters (therefore improving disaster resilience)
A treasure of data and possible finding currently beyond our imagination.
Thank youContact:
George Mitri, Ph.D.
Director of the Lannd and Natural Resources
Program, Institute of the Environment, University of
Balamand (BP-IOE-UOB)
Associate Professor, Faculty of Arts and Sciences,
UOB
Email: [email protected]
University of Balamand
P.O.Box: 100, Tripoli - Lebanon
Tel: 00961-6-930250 ext. 3944
Fax: 00961-6-930 257
www.balamand.edu.lb
Other links:
home.balamand.edu.lb/wildfire
Thank you