20190214 Presentation to Lloyd'sIntelligence+Presentation+-+MIS+Intel.… · and associated press video imagery feeds. Analysts used the inaccurate FEMA data and the video feeds to

www.MIS-Intel.com

Forbes McKenzie

www.mckenzieintelligence.co.uk

McKenzie Intelligence

Forbes Mckenzie

Scope


● What is the MIS-Intel portal?

● What does it do?

● Overview of Reports & Data

● Using MIS-Intel

● Live Demo

● Good to know!

● Analyst’s Damage Assessment (extra)


Using MIS-Intel


Home Page

Reports Page Example Report – Hurricane Michael

Login & Password change

What does it do?McKenzie Intelligence

● Situational Awareness.● Pre/Post Imagery.● Search facility● Aids decision making.

Exposure Report (24hrs)


● ZIP Code or Grid level

● Damage Assessment

● No imagery - OSINT

● All types of NATCAT events

Exposure Report (24hrs) – OSINT = IoTMcKenzie Intelligence

Exposure Report (24hrs)McKenzie Intelligence

Exposure Report (24hrs)McKenzie Intelligence

Claims Report (48hrs)McKenzie Intelligence

● Before/After imagery slider

Timeline ReportMcKenzie Intelligence

Hurricane HARVEY: Houston - Flooding Event

Background:

After making landfall on 25 August 2017, as a Category 4 storm, Hurricane HARVEY, quickly decreased in both intensity and momentum.

HARVEY in effect stalled over the low lying Texan coastline and the warm waters of the Gulf of Mexico for 6 days. As such the whole of south-eastern Texas and south Louisiana suffered persistent and intense rainfall as the huge weather front slowly progressed inland towards Tennessee.

Extensive, flash flooding occurred within the major conurbation of Houston; purportedly permanently ‘sinking’ the city area by 0.05 m.

© MIS Ltd. (2017) Imagery & Mapping © Digitalglobe Inc (2017) Weather Radar © NHC/NWS/NOAA (2017)


Landfall 0 - 24 Hours: Houston

As soon as Hurricane HARVEYs’ US landfall was classified as a NATCAT by the Lloyd’s Market, MIS Imagery Analysts aided by the All-Source Research team conducted first line 24 Hour reporting.

Despite a lack of coherent information; a fluid large scale event; a lack of ground truth and a realisation the FEMA modelled data was inaccurate the 24 Hour report provided at 10 kmsq resolution:

• The level of Market Exposure for Senior Decision Makers at scale.

• Accurate, reliable, geospatial data; fused information and textual reporting.

Key Recommendations pertaining to risk and asset loss.

© MIS Ltd. (2017) Imagery © Digitalglobe Inc (2017) FEMA Probability Model at 1.5 NM scale © FEMA (2017) Analytics Data © MIS Ltd (2017)


FEMA & State Modelled Data:

The slow nature of HARVEYs’ momentum prevented MIS fully utilising satellite imagery collection, due to extensive thick cloud cover.

MIS geospatial analysts quickly recognised that the unique event would render Federal Emergency (FEMA) flood and flow models inaccurate.

Despite this, MIS were able to provide large scale, by zip code analysis of flooding in Houston using:

• Social media and images from CCTV

• State River gauge monitors

• Synthetic Aperture Radar

© MIS Ltd. (2017) Imagery & Mapping © Digitalglobe Inc (2017). FEMA Model © FEMA (2017) Analytics Data © MIS Ltd (2017)


6 - 48 Hours: Social Media &

Video Data

MIS provided the market at the up until the

48 hour point with an overview of Houston

flooding at zip code level. This was based

upon non- traditional analysis of social, city

and associated press video imagery feeds.

Analysts used the inaccurate FEMA data and

the video feeds to attempt to assign

probability of damage or flooding based

upon all the available information. This

information provided some early insight, but

a lack of full coverage precluded full city with

analysis using these data sources.

© MIS Ltd. (2017) Imagery & Mapping © Digitalglobe Inc (2017). Analytics Data © MIS Ltd (2017)

© MIS Ltd. (2017) Generic Houston CCTV system.


24 -72 Hours: River Gauges

MIS provided more veracity and clarity to the single sourced video data through the monitoring of State / National NOAA / NWS river flow and high water gauges.

The gauges provided near real time ground truth geospatial data. This data enabled MIS analysts to plot and correlate the river flow and flood status data. Furthermore, analysts were able to plot down to the minute when rivers peaked and overtopped their banks.

MIS analysts were then able to confirm or deny uncorroborated often single direction facing video information streams. Armed with this dual sourced data in conjunction with FEMA modelled data MIS provided early flood damage insight and levels of exposure at zip code level to the Lloyd’s Market.

© MIS Ltd. (2017) Imagery & Mapping © Digitalglobe Inc (2017) NOAA Data © NWS/NOAA (2017) Analytics Data © MIS Ltd (2017)


72+ Hours: SAR (Radar Data)

MIS utilised the properties of low resolution

(20m) Synthetic Aperture Radar (SAR) to

map at scale areas of standing flood water.

SAR utilities the differences in radio wave

returns to provide and build a low resolution

ground picture. MIS analysed and

contrasted pre event SAR with post event

SAR. The differences in the SAR returns

were separated into a single vector layer.

This layer represented the flood waters;

albeit as a series of 20m x 20m pixels.

© MIS Ltd. (2017) Imagery © Digitalglobe Inc (2017) & Copernicus Sentinel Data © ESA (2017)

Analytics Data © MIS Ltd (2017)

© MIS Ltd. (2017) © Digitalglobe Inc (2017) Copernicus Sentinel Data © ESA (2017)


7 - 10 Days: Aircraft Imagery

As Hurricane HARVEY dissipated US Federal

authorities began overflights of areas of

flood within the wider Houston environs.

Very high resolution (less than 0.5m)

imagery swathes were imaged from low

altitude. This imagery further added

credence to MIS all-source and non-

traditional imagery information

© MIS Ltd. (2017) NOAA Imagery© NOAA (2017)

Imagery & Mapping © Digitalglobe Inc (2017) Analytics Data © MIS Ltd (2017)


Geoeye 1

Geoeye 2

Quickbird

Worldview 1

Worldview 2

Worldview 3

HiRes Satellite Imagery

Bespoke TaskingMcKenzie Intelligence

Bespoke TaskingMcKenzie Intelligence


Demo

How do I get access?McKenzie Intelligence

● Send email to [email protected]

● A password will be sent to your email address

mailto:[email protected]

MIS-Intel – Forward vision – Increased automation in order to

decrease property assessments at scale


● Exploitation of increased availability of data (IoT)

● Increase in automation (machine learning)

● Ingestion of client data (Blockchain)

● Creation of Authoritative damage layer

● Application Programming Interface (API)

Ingesting Risk DataMcKenzie Intelligence


API Development


Visualisation of Damage


Any Questions

20190214 Presentation to Lloyd'sIntelligence+Presentation+-+MIS+Intel.… · and associated press video imagery feeds. Analysts used the inaccurate FEMA data and the video feeds to

Documents