CDRP – Climate Disaster Recovery Process In Chennai

CDRP – ClimateDisaster RecoveryProcess in ChennaiSummary ReportZurich, Switzerland, December 2017University of Madras

About the InitiativeThe Climate Disaster Recovery Process Initiative for Chennai is an umbrella project of the ClimatePolicy group of ETH Zurich. This initiative builds on earlier work conducted between 2009 and 2012as part of the Climate Disaster Resilience Initiative. This current research project was developedin cooperation with the University of Madras and with support from the Greater Chennai Corporation, the Revenue Administration of the Government of Tamil Nadu, MLA offices of Mylapore andVelachery constituencies. A full master thesis of this research project is available upon request.TitleCDRP – Climate Disaster Recovery Process in Chennai: Final Summary ReportReport prepared byDr. Jonas Joerin1Prof. Ramasamy R. Krishnamurthy2Franziska Steinberger1Dr. Anna Scolobig1AcknowledgmentsDr. K. Satyagopal, IAS, Principal Secretary/Commissioner of the RevenueAdministration of the Government of Tamil NaduDr. D. Karthikeyan, IAS, Commissioner of the Greater Chennai CorporationDr. R. Nataraj, IPS, MLA of Mylapore ConstituencyMr. Vagai Chandrasekhar, MLA of Velachery ConstituencySpecial thanks to various representatives from residential welfare associations, NGOs and youth organisations to support data collection andvalidation process.Title pictureNew settlements in Perungudi, ChennaiLayoutSandro Bösch 2017 ETH Zurich1 ETH Zurich, Department of Environmental Systems Science, Universitaetstrasse 16, 8092 Zurich, Switzerlandjonas.joerin@usys.ethz.ch, 41 44 632 32 402 University of Madras, Department of Applied Geology, Guindy Campus, Chennai 600 025, Indiarrkrishnamurthy@gmail.com, 91 9677045519

1Summary ReportExecutive SummaryAfter an unusual period of prolonged intensive rainfall caused by a coincidence ofthe northeast monsoon (October-December)and the El Niño, Chennai was badly affectedby the 2015 South Indian floods. As Chennaiis built around a former lagoon andis thus located at the end of a series of water bodies, itis vulnerable to excessive periods of rainfall.Since the catchment areas in KanchipuramDistrict were already filled-up by the end ofNovember 2015, the relentless downpour ofrainfallon December 1, 2015, triggered thefloods and forced the city to stand still. Onthis day in average and citywide 286mm ofrain fell within 24 hours[1] and the catchmentareas of Adyar and Cooum received about 490mm of rain. The amount of rain recorded onDecember 1 was the highest in 100 years[2].Damages were estimated at USD 3 billion and301 casualties were recorded in Chennai.In response to this disaster, we assessedthe recovery process 10 months after througha household survey in two equally exposed(coastal) constituencies, Mylapore and Velachery. The objective was to understand howthe physical, social and economic conditionsof the city recovered following this disasterevent. Furthermore, we looked at the interplay between communities and authorities inthe recovery process. Following the household survey, survey validation meetings wereheld with residents of Velachery (June 29,2017) and Mylapore (June 30, 2017).Key findings from this study include: Residents living in low-land (marshland) and newly built areas in Velacherywere more affected by damages (houses,household assets, health) compared toresidents living in the older part of Chennai (Mylapore). The median damages perhousehold wereINR 30,000 in Mylaporeand INR 50,000 in Velachery. The recovery time (average) for basic services (electricity, water, sanitation, solidwaste management, roads and communication) took about seven days in Mylapore and around 14 to 21 days in Velachery.Housing took in average 22 days in Mylapore and 33 days in Velachery to be rebuilt. The recovery time for socio-economic factors (physical health, mental health, employment, household income, education,household assets, etc.) took longer to berestored compared to the physical items.For example, mental health took in average up to four months to be recovered andhousehold income losses more than threemonths in both areas. While physical items took longer to berecovered in the more affected area(Velachery), socio-economic factors tookequally long to be restored in Mylaporeand Velachery. The level of satisfaction about the recoveryprocess did not differ between residentsliving in Mylapore compared to Velachery.Therefore, the recovery time does not correlate with the level of disaster affectedness. Residents in the more affected area (Velachery) became more active to be betterprepared (stock emergency supply, maketheir home flood proof, get informed aboutflood mitigation options, etc.) for a futureflood disaster compared to residents inthe less affected area (Mylapore). Residents in the more affected area (Velachery) also become more solidary compared to residents in the less affected area(Mylapore). For example, more residentsparticipate in a volunteer group, talk toother members of their community andrequest assistance from neighbours. Residents in both areas showed equallyhigh willingness (around 90%) to becomepart of the recovery planning process. The absence of a needs assessment andlimited involvement of residents in therecovery planning process are key areaswhere the interplay between residents andauthorities needs to be improved in the future. Overall, this study summarises that theefforts by the various governmental agencies were effective in restoring physical,social and economic items.

2CDRP – Climate Disaster Recovery Process in ChennaiForeword message

3Summary ReportIntroductionAfter an unusual period of prolonged intensive rainfall caused by the coincidenceof the north-east monsoon (October-December) and the El Niño, Chennai was badlyaffected by the 2015 South Indian floods. Arelentless downpour of rainfall forced thecity to stand still on December 1, 2015 afterin average and citywide 286mm of rain fellwithin 24 hours [1]. Damages were estimated at USD 3 billion and 301 casualties wererecorded in Chennai. The airport and basicservices stopped functioning, e.g. communication lines and electricity were interrupted.Figure 1 highlights how the 2015 South Indian floods unfolded and affected Chennai.The 2015 South IndianFloods in ChennaiFigure 1Timeline of the 2015 South Indian floods in ChennaiThe severity and unprecedentedness ofsuch an extreme flood event in Chennai,called for a thorough assessment to understand to what extent the city could recoverfrom this disaster. Hence, the key objectivewas to better understand the physical, socialand economic drivers of recovery following aflood disaster event. Furthermore, we lookedat the interplay between communities andauthorities in the recovery process.Goals of study

4CDRP – Climate Disaster Recovery Process in ChennaiMethodologyFew studies exist that quantitatively assess disaster recovery processes [3, 4].Hence, this study aims to contribute to theunderstanding of the factors driving recoveryprocesses. For this purpose, we selected twoadjacent constituencies (Figure 2) with equalcoastal exposure and inclusion of waterbodies (e.g. Adyar river, canals in Mylaporeand basins/lakes in Velachery). However,Mylapore forms part of the old town of Chennai whereas Velachery was just developedduring the last two decades and containsmarshlands. The population of Mylapore isabout 300,000 and 600,000 in Velachery.We identified households in both areasthrough a stratified random sampling process. In total, 257 households in Mylaporeand 264 households in Velachery were surveyed during October and November 2016.The content of the survey included questionsthat are typical to a post-disaster needs assessment, such as the provision of basic services (e.g. electricity, water, sanitation, housing, etc.) and socio-economic factors, suchas household income, employment, household assets, education, health, nutrition, etc.For all these factors, we asked householdshow long they needed to recover and whetherthe services/factors were better, the same,or worse than before. Additionally, we askedresidents whether they became more active,solidary and engaged in their communitiesfollowing this disaster.MylaporeBay of BengalConstituencyRiversVelachery5kmNFigure 2Study area: Mylapore and VelacheryThe analysis of the survey focused oncomparing the responses between the twoconstituencies through statistical tests, including Mann-Whitney-Wilcoxon tests. Descriptive analysis, including boxplot analysis,complemented the tools on how to extractrelevant information from the survey data.Validation meetings with residents from theaffected constituencies served to verify theanalysed results.

5Summary ReportResultsAmong the surveyed households, the median of the total recorded damages (damageson house, household assets and health) wasINR 30,000 in Mylapore and INR 50,000 inVelachery. As Figure 3 shows, around 75% ofthe households in Mylapore recorded damages below INR 60,000. In contrast, the sameproportion (75% of the survey respondents)of households in Velachery recorded damages up to INR 100,000. This highlights thathouseholds in Velachery were considerably more affected by the 2015 South Indianfloods compared to Mylapore.Figure 3Total median damages in Mylapore and VelacheryThe reason for this is largely because thefloods were higher in Velachery (as high as3m above the ground) compared to Mylapore.A lack of effective drainage systems, particularly in Velachery, is one of the key reasonswhy the water could not run off. Furthermore, Velachery is located in a low-lying areawith marshlands. The damage structure(Figure 4) shows that a majority of the damages was attributed to damages on housesand a slightly smaller fraction was recordedon destroyed household assets.A somewhat weak correlation was foundwith the Pearson’s product-moment correlation for the variables ‘annual income’ and‘total damage costs’ (p 0.006, r 0.136), indicating that households with higher annualincome also suffered from higher damages.Moreover, Kendall’s tau correlation betweenthe variables ‘education’ and ‘annual income’(p 2.42e-14, tau 0.298) show (weak correlation) that households with higher annualincome also tend to have higher education.A significant correlation between the lev-Damages on households

6CDRP – Climate Disaster Recovery Process in Chennaiel of ‘education’ and ‘total damage costs’could however not be found. This highlightsthat the level of education is not decisive inwhether households are affected (householddamage-wise) by a natural hazard, such asa flood. Mylapore – compared to Velachery- has a higher proportion of non-educated100%90%80%70%people (32.27% in Mylapore and 16.87% inVelachery) and also a lower proportion of college-educated people (24.30% in Mylaporeand 40.16% in Velachery). This shows thatthe level of education did not prevent households from being affected by the 10%0%MylaporeHousingVelacheryHousehold assetsHealthFigure 4Damage structure (median values in INR) in Mylapore and VelacheryTherefore, the main reason for whyhouseholds were affected must be attributed to their actual exposure to the floods.The fact that the damage structure (Figure4) does not vary between the two constituencies confirms that the level of education cannot serve as an explanation for why households suffered from damages on their house,household assets and health.The offer by the Tamil Nadu State government for damage compensation was collected by 75% of the surveyed households.Among the 25% who did not receive damagecompensation, half responded that they didnot need it and the other half said that theydid not receive it.