Flooding & water availability in India

[tdobermann]

People: Robin @robinslburgess , Azhar @azharhsain , Nikita @NikitaSharma30 , Tim

Strand I: examining how flood events impact economic outcomes and modelling how flood exposure may change over time Task 1: establish a relationship between historical flood events and economic outcomes

• Compile list of datasets that document historical flood events across India (variables, geographic scope, precision/resolution, and time period)
• Compile list of economic datasets that measure outcome variables of interest (agricultural yields, employment, sector shares, capital, etc).
  • Indicate lowest geographic unit we can measure these outcomes on (e.g. district, village, individual, GPS)
  • Also add in satellite data (e.g. night lights)
• Begin merging datasets to start simple event-study analysis to see if we are picking up any meaningful effects on economic outcomes

Task 2: use hydrological models to forecast how planned precipitation changes, glacial melt, etc. influence river flow and subsequently flood exposure in the future

• Determine main inputs (variables like river flow) for core hydrological models to assess data requirements for forecasts
• Use historical data to test the first stage: does the model well predict flood events which we have seen in the past?
• Apply to a standard climate model which includes changes in precipitation etc

Strand II: examining how surface water availability is expected to change over time and what the economic impacts of this will be Task 3: closely linked to Task 2, list the main hydrological models used for determining water availability for a given cell in a grid

• List the data inputs that go into these models (e.g. river flow, precipitation, land use, terrain, etc)
• Using similar economic data as in Task 1, explore whether the second stage is interesting: can we show how changes in surface water availability lead to meaningful changes in economic outcomes?

Task 4: forecast how climate change will lead to changes in water availability

[tdobermann]

Data sources - Flooding.docx

[tdobermann]

Meeting: February 5, 2021

India WRIS

• Flood data, rainfall data all marked in maps. Micro data is mostly logged as classified
• Establishing an account does not work.. page does not proceed
• Email WB Anju Gaur
• Nikita to email WRIS platform about setting up an account

Darmouth flood data

• Now being hosted by UColorado with larger database
• 1985 onwards database
• River gauge for only the first decade (2002-2009)

NFHS

• 2005 no GPS, GPS only 2014/15
• More health focused, family planning

NSS

• Microdata catalogue
• More consumption focus
• Look more into dates and geographic targeting next week
• Azhar to send details / links to Nikita to access NSS
• Not stored in a convenient format

Firm datasets

• ASI: explore sub district
• Prowess (good for learning about how workers are organised in a firm)

Village maps

• Azhar to send what info is available on village level maps

[tdobermann]

Meeting: February 26, 2021

• Hydrology chat: Tim & Azhar to do further homework and then we convene a session later to go through planned details on model
  • Post notes on Git from papers; coordinate the set of papers we are reviewing
• Preliminary analysis on the datasets: use NSS to do some basic descriptive analysis
  • Start with 68th NSS rounds
  • Summary statistics of consumption at the district level (mean consumption for instance)
  • Also look into agricultural variables
• Supreet Kaur nominal wage rigidity data:
  • World Bank agricultural dataset unclear if accessible
  • UDel rainfall dataset; GMFD climate data maybe more preferred
  • Compiled entire NSS dataset across all rounds
• Tim to look into further agricultural data