Monsoon flooding in North-East India: Identifying those in need using satellite data

The current floods in the state of Assam, northeast India, are the worst in a decade and around 2 million have had to leave their homes ( However, Assam experiences monsoon floods every year and these are vital to the state’s rice agriculture as the majority of land is not irrigated. In extreme years, when the floods are much larger than usual, as they have been this year, it can cause huge problems for the state. The extreme flood events can lead to the loss of life, property and livelihoods that are being reported in the media. However, the floods also cause huge problems for the longer term development of the state. There is evidence to suggest that flood events in northeast India are increasing in frequency, with economic costs rising and limited government resources that means only a fraction of the damage can be repaired using government support each year. For example, the World Bank has estimated that annual flood damage is around US$163 million in Assam but the state only receives approximately US$21 million from the central government (World Bank 2007).

Living in poverty

Despite India having one of the fastest growing economies in the world, there have been large variations in the rate of development of different states within India.  Assam has struggled to keep up with the rest of India.  The lack of development in Assam has led to persistently high levels of poverty.  Development in Assam has been low and poverty rates relatively high partly due to a large dependence on agriculture for incomes and subsistence.  Thus, extreme floods mean that large numbers of people lose their only incomes and source of food.

Can money and aid help?

The limited financial support available is not just a feature of India.  All governments have limited financial capacities.  Thus, extreme climatic events (flooding in the Assam case but could be drought, tsunamis etc. in other locations) coupled with rapidly changing social and economic conditions mean it can be difficult to direct emergency aid and development assistance to those most in need.  The Air Force in Assam is currently supplying food parcels to those stranded by the floods (  However, it is unlikely that all the aid will reach those most in need and this is where targeting of resources becomes useful.  Data on social and economic conditions such as those provided by the Census (national survey) can be used to identify those most in need and  these facts can in turn by used to  inform assistance programmes to ensure that aid is delivered efficiently. However, these surveys are expensive to conduct and can take time to update.  This is where images from satellites may become very useful in the future.

The value of data from space

My research over the last 4 years has aimed to predict social or economic (socioeconomic) conditions within villages by comparing local land cover characteristics identified from only satellite data to the known literacy rates measured in the 2001 national census, which can be used as a proxy for poverty.  It was based on the idea that correlations are often found between land cover and poverty/development.  Since satellites can be used to classify different land cover types it is logical that the land cover information provided by satellite data may also provide us with some information on the socioeconomic conditions of communities.  The research first identified links that had been found between poverty and land in studies by other researchers.  Then, satellite data was used to identify specific land cover types (e.g. woodland, water etc.) that had known links with poverty.

The proportion of different land cover types within a village and its surrounding area were estimated and statistical models were used to find out if local socioeconomic conditions could be inferred from this land cover information.  The socioeconomic conditions chosen were:

  • Female literacy because it was often higher in villages with wealthier conditions such as access to domestic electricity and permanent modern housing construction, and;
  • Non-agricultural employment because it was often higher in villages with wealthier conditions such as access to domestic electricity, permanent modern housing construction and access to personal motorised transport.

The link between poverty and land cover

The results indicated that significant associations exist between female literacy and land cover types, and non-agricultural employment work and land cover types. Many of the associations identified could be interpreted meaningfully in relation to both the understanding gained from data collected when visiting Assam and in relation to generally accepted associations from other researchers’ findings. Table 1 and Table 2 provide some examples as to how the land cover relates to socioeconomic conditions. Although the results are difficult to compare directly with studies that have explored the associations between poverty or economic development and the environment, the quantitative findings of the research were in keeping with expectations and research hypotheses, lending credibility to the associations observed by other researchers.

Table 1: How female literacy is associated with a range of remotely sensed environmental variables in Assam.

Female Literacy is: When a village:
65% more likely Has more than 50% land area covered by Woodland
54% more likely Has more than 40% land are area covered by summer crop
10% more likely Is within 1km of the average annual flood zone (but not inside it)
37% less likely Has more than 20% land area covered by winter cropland
26% less likely Is over 1km from a main road
19% less likely Is over 90 minutes from a major market town

Table 2: How non-agricultural employment is associated with a range of remotely sensed environmental variables in Assam.

Non-agricultural employment is: When a village:
59% more likely Is over 1km from the average flood zone
43% more likely Has over 60% land area covered by woodland
36% more likely Has a road within the village boundary
49% less likely Is over 10km from a main road
25% less likely Is over 10km or over 90 minutes travel time from a major market town
22% less likely Has over 20% land area covered by water.

Using satellite data research to assist with development in Assam

In the future, data from satellites (a.k.a “remote sensing”) could be used in conjunction with census data to monitor socioeconomic changes at more regular time intervals than are currently available. Census data cannot be replaced by remotely sensed environmental data as ground surveys provide information on non-environmental issues important for development that satellite data is incapable of providing. However, monitoring change in environmental factors in conjunction with data on socioeconomic conditions would potentially be useful for targeting resources to locations where development assistance is needed the most. For example, monitoring flood extents and overlaying predicted socioeconomic conditions onto this may help identify communities that are most vulnerable to descending into poverty as a result of the flood event.  Assistance can then be targeted in these areas.

more information can be found at my website


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: