Visualising natural disaster losses from CATDAT in HiDE

Last update: March 23, 2011 at 11:26 am by By

Simply click below to get started viewing natural disaster losses! The new version by Aidan Slingsby runs directly in the browser with no need for setup!

Click here to start running CATDAT-HiDE in browser

You will remember that CATDAT presented their annual reviews of Damaging Earthquakes and Damaging Volcanoes in 2010 through

CATDAT collates statistics on natural disasters and is looking for new ways to help people generate knowledge from its data and to disseminate it. Here, we are evaluating the approach that HiDE takes to facilitating data exploration and sharing findings using Twitter. Our example dataset includes 760 earthquake events, 200 volcanic events and 2700 flood events between 1988 and 2010.

HiDE is software that lets you construct information graphics using visual variables such as position, size and colour to encode data in different ways at different hierarchical levels. This makes various kinds of comparison possible.

HiDE integrates with Twitter enabling graphics to be shared along with short commentary about what the graphic tells us.

HiDE can import and export graphics using the HiVE language. HiVE is automatically exported with every tweet. It can be copied and pasted to and from HiDE and used as a means to record, share, archive and recreate graphics produced at different times and by different people.

This collaborative work between CATDAT,, CEDIM, giCentre (City University London) and the Willis Research Network is evaluating interactive visual approaches for data analysis.

How to take part

Taking part is easy:

  1. Download: the HiDE software and the CATDAT dataset [download]
  2. Explore: by building successive graphics to reveal structure, compare patterns and generate insights into Natural Disaster losses in the last 22 years.
  3. Tweet: save graphics and comments to record your exploration and share your insights. You may also like to post graphics and comments onto your blog.
  4. Search: Use the #viztweet and #catdat hashtags on Twitter to find out what others are saying [search]
  5. Discuss: Try out the graphics that others have build and respond to their comments.
  6. Feedback: Let us know how you got on by filling in our questionnaire.

Please watch our video which demonstrates how to use the software.

Example tweets

sHier(/,$peril,$year); sOrder(/,HIER,[NULL,HIER]); sSize(/,$count,$totalFatalities); sColor(/,HIER,$totalLoss); sLayout(/,SF,SF);

Number of fatalities, coloured by sum of loss, by peril and by year. Largest losses for flood were in years with few fatalities!

sHier(/,$peril,$month); sOrder(/,HIER,[NULL,HIER]); sSize(/,$totalFatalities,$totalFatalities); sColor(/,HIER,$totalLoss); sLayout(/,SF,SF);

Most earthquake-related deaths in Jan and Mar. More flood related deaths in Jun-Aug - probably seasonal.

sHier(/,$year,$country); sOrder(/,[HIER,NULL],[NULL,HIER]); sSize(/,$totalFatalities,$totalFatalities); sColor(/,HIER,HIER); sLayout(/,SF,SF);

Fatalities by year and country across all perils. Usually concentrated in one country. 2004 was a particularly bad year dominated by the Indonesian Tsunami.

sHier(/,$peril,$year,$country); sOrder(/,HIER,[HIER,NULL],[NULL,HIER]); sSize(/,FX,$totalFatalities,$totalFatalities); sColor(/,HIER,HIER,HIER); sLayout(/,SF,SF,SF); sFocus(/FLOOD/,SL);

As before for flood only. Dominated by a 1991 Bangladeshi flood and the 2008 Burmese Cyclone Nargis - causing widespread flooding.

Why we are doing this

More and more datasets are being released to the public and visualization is being used to make these data more accessible to people.

The giCentre is interested in the process of using graphics to explore data and how insights from this process can be described, archived and reported. HiDE was designed to help with this and would like to evaluate the informed reactions of a motivated group of users visualizing an interesting data set.
The Willis Research Network funds collaboration between academia and the insurance industry and promotes the application of the latest research to the needs of the global insurance and reinsurance industry.

The data

Our example dataset includes 760 earthquake events, 200 volcanic events and 2700 flood events between 1988 and 2010. Only major events* are included, representing only a subset of the full database. Details of each event include deaths, injured, displaced, and 2010-adjusted economic losses. Values have been rounded.

For full details of CATDAT parameters please go to, and download the CATDAT Annual Reviews of 2010.

* Earthquakes: at least 5 dead, 5 injured or $100m. Volcanoes: at least 1 death, 1 injured, 1000 evacuated, or $10m. Floods: at least 5 dead, 5000 displaced or $100m. Costs are in Country HNDECI Adjusted 2010 US-Dollars.

The Disclaimer

Any use or quoting of this data must be referenced containing the author "James E. Daniell" and the keyword "CATDAT". References to other CATDAT reports are also acceptable for the keyword.


CATDAT originated as a series of databases that have been collected by the author and founder, James Daniell, from many sources over the years (2003 onwards). It includes global data on floods, volcanoes and earthquakes (and associated effects).

The databases were presented at the Australian Earthquake Engineering Society Conference in 2010 in Perth, Australia, in the form of 3 papers, and the data was also used to form an Asia-Pacific comparison of flood and earthquake socio-economic loss in the CECAR5 conference in Sydney, Australia, 2010. In addition, many new and exciting trends have been established allowing for in-depth insights into socio-economic loss from natural disasters using the global socio-economic indicator work of Daniell (2008-2011).

As of February 2010 in CATDAT Damaging Earthquakes Database v5.01, over 17000 sources of information have been utilised to present data from over 12200 damaging earthquakes historically, with over 7000 earthquakes since 1900 examined and validated before insertion into the worldwide CATDAT damaging earthquakes database. In CATDAT Damaging Volcanoes Database v2.82, over 2500 sources of information have been utilised to present data from over 1650 damaging volcanic incidents historically, with over 950 volcanic incidents since 1900 examined and validated before insertion into the CATDAT Damaging Volcanoes Database. In addition a subset of the CATDAT Damaging Floods Database v2.10, is also presented with about 3750 worldwide flood incidents from 1988 to 2010.

CATDAT currently presents its information through, the SOS Earthquake website (founded by Armand Vervaeck), and CEDIM (Center for Disaster Management and Risk Reduction Technology, Karlsruhe, Germany).


  1. I originally had problems with running the hide file. It may be the case that the program does not pick up where java is....
    Thus just open up HiDE.bat, and redirect to where java is..
    Usually C:\Program Files\Java\jre6\bin\java

Desktop Version