A District level Flood Severity Index for India

Manabendra Saharia1, Sharad K Jain2, Ved Prakash1, Harshul Malik1, O P Sreejith4

1
Department of Civil Engineering, Indian Institute of Technology Delhi, Hauz Khas, New
Delhi 110016, India
2
Department of Civil Engineering, Indian Institute of Technology Roorkee, Roorkee,
Uttarakhand 247667, India
3
India Meteorological Department, Pune, India

Corresponding Author:
Manabendra Saharia, Ph.D.
Assistant Professor, Department of Civil Engineering
Associate Faculty, Yardi School of Artificial Intelligence
Indian Institute of Technology Delhi
New Delhi, India 110016
HydroSense Lab: https://hydrosense.iitd.ac.in/

1
Abstract
India is one of the worst affected countries in the world in terms of fatalities and economic
damage due to natural disasters, particularly floods. For planning flood mitigating and relief
measures, granular historical information on a pan-India basis is required, which has been
missing. Through recent efforts, a few national scale datasets have been created, but they lack
the requisite information on fatalities and damages, which has limited the ability to develop a
flood severity index. This paper describes the development of the India Flood Inventory with
Impacts (IFI-Impacts) database, which contains death and damage statistics, and combines
population and historically flooded area information sourced from a national hydrologic-
hydrodynamic modeling system. We also propose a novel District Flood Severity Index
(DFSI), which accounts for the historical severity of floods in India based on the number of
people they have affected and the spread and duration of such floods. Districts being the
administrative units of the government, this novel index fulfills a major need and gap in
currently available flood management tools. The dataset as well as the index is expected to
significantly advance disaster preparedness towards floods in the country. DFSI can be
improved further by collecting and incorporating additional variables, e.g., economic losses
and by improving the reliability/robustness of the data of other variables. Based on DFSI,
actions need to be addressed to mitigate flood damages, beginning with the districts with the
high DFSI values.

Key words: flood database, inundation, population, hazard assessment, flood severity index,
flashiness, ILDAS, DFSI

Highlights (3-5 bullet points that capture the novel results of your research as well as new
methods that were used. Be clear and concise and use simple terms)
1. India faces floods recurrently, causing huge adverse impacts to life and assets.
However, a comprehensive flood severity index is missing.
2. A district level flood severity index is developed using the India Flood Inventory with
Impacts (IFI-Impacts) database containing data of variables representing the
occurrence of floods and damage due to floods.
3. Since district is the most relevant unit for planning and decision making, a DFSI would
be of immense value for flood management.

2
 1. Introduction
India is subjected to monsoonic climate in which about 75% of the annual rainfall takes place
over the rainy season of four months (June-Sep). A large concentration of rainfall leads to
recurrent floods in many parts of the country. The impacts of floods in India have been severe,
with 113,390 human casualties reported between 1975 and 2015, averaging 2765 deaths per
year (Saharia et al., 2021a). In addition, floods also damage infrastructure including bridges
and roads, crops, building, etc., inflicting economic losses. Many efforts are being made in
India to manage floods so as to protect human lives and property. To design better flood
management strategies and focus attention to the most critical areas, an essential requirement
is to know the places that are most vulnerable to flood damage. A flood severity map captures
the spatial and temporal information in terms of extent, duration, and damage caused by floods.
While flood hazard studies over India using hydrologic or statistical models are abundant in
literature, no flood severity index map that incorporates ground level information at fine scales
is currently available.
Preparation of flood severity map for a large and diverse country like India is
challenging due to several reasons. Data related to floods belongs to a number of sectors:
meteorology, hydrology, agriculture, transportation, urban, rural, revenue, health, disaster
management, etc. Floods data are collected by a large number of agencies with different
objectives, mandates, skills, and available tools. The formats of data collection are also widely
varying. Currently, there is no national/regional scale system to compile the data which may be
buried in paper records, computer files, maps, etc. Often, there is lack of data sharing and after
the event or intended use/application of the data is over, it finds very little usage in long-term
flood management studies.
Studies show that the spatial organization of rainfall influences flood severity at par
with geomorphology and climatology (Saharia et al., 2021b). India being a large country, wide
variation in these causative factors is seen. In the catchments in north India (Ganga,
Brahmaputra, and Indus), floods may occur due to intense rains as well as snow/glacier melt.
In recent times, frequency of occurrence of Glacial Lake Outburst Floods (GLOFs) is rising;
specifically, three GLOF events (Sikkim 2023, Rishiganga 2022, and Kedarnath 2013) have
caused large number of deaths, and damage to infrastructure, roads, bridges, and buildings.
Moreover, frequency of high intensity rainfall events is also rising, leading to severe flooding
at many places. Besides the Himalayan river basins, floods also occur in the catchments of

3
Godavari, Narmada, Mahanadi, Tapi rivers, and in the state of Kerala. Apart from these,
frequent urban floods (Mumbai, Chennai, etc.) have also caused much damages and have
impacted the economy.
To assess myriad impacts of floods, we require datasets that not only capture the
information about the floods, but also on the consequent losses and damages. While there are
national scale flood modelling systems such as the Indian Land Data Assimilation System
(Magotra et al., 2024), there is a growing need for flood severity indices that are based on
observational datasets. Various flood severity indices have been developed for the United States
(Saharia et al., 2017; Schroeder et al., 2016), Europe (Marchi et al., 2010), Greece (Diakakis et
al., 2020). But in India, such studies have been limited to case studies (Ghosh and Dey, 2021;
Kanth et al., 2022), and a national flood severity index has been missing. The risk of floods at
a given place depends on the climate of the region, properties of precipitation upstream of the
place, topography, soils and geology, land use/cover, and human factors/interventions.
Concerned about the flood occurrences in India, Government of India had constituted
the National Commission on Floods (Rashtriya Barh Ayog, RBA) in 1976 which submitted a
comprehensive report in 1980 (Rashtriya Barh Ayog, 1980). Till date, this report is one of the
most comprehensive attempts to study the problem of floods in India and continues to be used
for various studies. RBA has put forth a number of recommendations and in the past efforts
have been made to implement these. Study of floods figures prominently in the activities of the
Central Water Commission (CWC), a technical organization under the Ministry of Jal Shakti,
and of the Ganga Flood Control Commission, Govt. of India. CWC also operates a Flood
Management & Border Areas Programme (FMBAP). In addition, different State Governments
also have the ministries that deal with flood management. National Remote Sensing Centre
(NRSC), an institute of the Indian Space Research Organisation (IRSO), publishes flood
inundation maps for different years prepared using satellite data. A number of strategies have
been suggested for flood mitigation in India. A view is that flood waters may be considered as
a resource to be conserved, to the extent feasible, and be used to meet the water demands in the
dry summer season which comes a few months after flood season (Jain and Singh, 2022).
CWC had published a map in 2023 on flood prone areas of India which is based on
slightly old data. Our analysis shows that the flood prone area in India has changed considerably
in the recent times and this map needs to be updated. In the last few decades, many floods have
occurred at places that are not considered flood prone. Recently, National Remote Sensing
Centre (NRSC 2023) has released a “Flood Affected Area Atlas of India - Satellite based
Study”, covering period of 25 years (1998-2022) by using both optical and microwave satellite

4
data. This atlas is closer to ground reality but also has inherent limitations as the estimated flood
extent depends on the availability of satellite data, date of overpass of the satellites and available
coverage over flooded areas. The NRSC atlas has presented data and analysis statewise. It
describes major flood events, district-wise flood affected area, and submerged roads and
railway. Optical and microwave satellite produced images are also included. The NRSC atlas
doesn’t cover flood events that lasted for short time and could not be mapped due to non-
availability of satellite data in short duration. Thus, it is expected that there may be
underestimations in flood affected area in the NRSC atlas. Further, the atlas does not provide
information about lives lost and economic damages, etc. Fig. 1 shows the CWC (2013)
produced map of area liable to flood and flood affected area map produced by NRSC. It is
noted that the two maps have some agreement and some differences. The area liable to flood is
more than the flood affected area.

Figure 1: (a) Map of flood prone area in India as identified by the Central Water Commission (CWC, 2013), and (b) Map of
flood affected area using satellite data from 1998-2022 by the National Remote Sensing Center, ISRO.

Despite the fact that flooding is a highly damaging disaster in India, there is a lack of
comprehensive data pertaining to floods or a dedicated flood portal in India. One such effort is
the building of the India Flood Inventory (IFI), which is the first freely available, analysis-ready
geospatial dataset over the region with detailed qualitative and quantitative information
regarding floods, including spatial extents (Saharia et al., 2021a). To draw plans for flood
management, data on many other features, related to flood damages are required at finer spatial
resolution. These data include, human causality due to floods, deaths of cattle, economic
damages, e.g., damage to crops, roads, bridges, buildings, infrastructure, etc. Some of these data

5
are collected by various government agencies but remain buried in official records, thereby
limiting their use. Researchers and academicians may not be aware of the existence of such data
and find it hard to access and use them. Further, well-articulated protocols to collect, validate,
and report these data have not been developed/followed.
This study presents an updated India Flood Inventory with Impacts (IFI-Impacts)
dataset, which contains detailed death, injury, damage and flood information from 1967-2023.
We also ran a national hydrologic-hydrodynamic modeling system to find historically flooded
areas of the country. Finally, we have developed a novel district flood severity index (DFSI)
which is based on district wise information on flood duration, fatality, injury, flooded area, and
population.

2. Causes of Floods in India

In India, floods are generated due to many different causes: pluvial, fluvial,
snowmelt/GLOF, cyclones, and drainage congestion. Among these, pluvial or rain-induced
floods are the most common. Floods may be generated by intense rainfalls of short to medium
duration or moderate rainfalls of relatively long duration or over large areas. In both cases, an
area may receive high volume of water which frequently exceeds the capacity to drain out the
water, leading to the inundation of land. In mountainous areas having loose and exposed soils,
water also erodes soils and flood waves may carry substantial quantities of sediments. Drainage
congestion is another reason behind flooding in some parts of India, e.g., Bihar plains, and
occurs when the rivers in an area are unable to drain out all the incoming flows. This results in
the piling of water and inundation.
Large changes in land use and land cover (LULC) in an area may significantly change
its hydrologic response. When the LULC changes result in larger generation of overland flows,
the area may begin to face flooding. Similar impacts may be seen after deforestation, unwise
urbanization, construction in low lying areas, etc. Climate change is likely to significantly alter
the statistical properties of the hydro-meteorological inputs to a catchment and its hydrologic
response. Due to climate change, localized intense rainfall events are likely to rise in future and
so more flash floods can be expected. Under the warming climate, incidences of cyclones will
increase and coastal and nearby inland areas will face more instances of intense rainfalls and
floods. It is also likely that the areas that are not currently flood prone may begin to see flooding.

3. Data Used and Processing

a. India Flood Inventory with Impacts (IFI-Impacts)

6
The original IFI dataset contained information on flooding from 1967-2016, sourced from an
annual printed publication named “Disastrous Weather Events” (DWE) by the India
Meteorological Department (IMD), but has undergone extensive cleaning and adaptation to
make it amenable for computational research. But the dataset had no district information, and
the focus was flood events, rather than impacts in terms of loss and damage to life and property.
The new IFI-Impacts database digitizes voluminous amount of information stuck in printed
documents published by various government departments in India, which have seldom been
used in research. This data is collected by IMD offices around the country and are ground-
validated, which increases the trustworthiness of the data in terms of ascertaining damages,
fatalities, as well as spatial extents. Apart from loss and damage data, the IFI-Impacts dataset
also contains the historically flooded area of a district in percentage derived from the
streamflow reanalysis provided by the Indian Land Data Assimilation System (ILDAS,
(Magotra et al., 2024)). Districts are the basic administrative units in India. But, some of these
have been reorganized and/or split over the time. In some instances, district names have been
changed and this posed a challenge in reconciling the data. Another difficulty was that in many
cases, more than one spelling of district name is used. Therefore, preparation of this database
required careful screening of data and manual corrections, which was a laborious task.

b. Flooded area from ILDAS

The Indian Land Data Assimilation System (ILDAS) was developed as a hydrologic-
hydrodynamic modeling system to estimate land surface states, channel discharge, and
floodplain inundation over South Asia (Magotra et al., 2024). It is built on top of the NASA
Land Information System Framework with supports multiple land surface models, data
assimilation schemes, and routing schemes (Kumar et al., 2006). Here, the Noah-MP land
surface model with multi-parameterization option was used with various physical processes
such as runoff generation, dynamic vegetation, canopy stomatal resistance, groundwater, and
so on (Niu et al., 2011). For discharge and floodplain inundation, we integrate Noah-MP with
the Hydrological Modeling and Analysis Platform (HyMAP; (Getirana et al., 2012)) river
routing model, which is a state-of-art global scale hydrodynamic model that simulates surface
water dynamics in rivers and floodplains using local inertia formulation. ILDAS verification,

ILDAS was run on a 0.1 degree spatial resolution at a 15 minute timestep with daily outputs
from 1981-2021. We use gridded daily rainfall data at 0.25° spatial resolution from the Indian
Meteorological Department (IMD). This gridded data is generated from 6955 gauge stations

7
using an inverse distance weighted interpolation scheme. The other model forcing variables
such as radiation and humidity were supplemented from Modern-Era Retrospective Analysis
for Research and Applications, Version 2 (MERRA-2; (Gelaro et al., 2017). The specifications
for various ILDAS parameters are given in Table 1.

Table 1: List of ILDAS components and their specifications

ILDAS Component Specifications

Land Surface Model Noah-MP 4.0.1

Routing Scheme HyMAP
Spatial Extent 68°-98°E, 5.5°-37.5°N
Spatial Resolution 0.1°
Temporal Resolution 15 minutes Noah-MP 3.6 and HyMAP with adaptive
timestep, daily output fields
Time Period 1981-2021
Forcing IMD, MERRA-2,
Forcing Variables Precipitation, near-surface air temperature, near-surface
specific humidity, surface pressure, eastward and
northward wind velocity, incident longwave and
shortwave radiation
Forcing Height 2 m for surface air temperature, specific humidity, and
surface pressure, 10 m for wind
Topography and river MERIT Hydro
network
Soils Definition (NCAR) STATSGO+FAO blended soil texture map
Vegetation Definition MODIS-IGBP (NCEP-modified), Monfreda et al. (2008)
crop types
Output Format NetCDF

Using a 40-year daily reanalysis data, we estimated the average flooded fraction of a district to
use as a robust but static descriptor of flood proneness of the district. To obtain the corrected
average flooded fraction and create the flood layer, we subtracted the permanent water layer
fraction from the estimated average flooded fraction obtained from ILDAS. The permanent
water fraction for each district was calculated from the permanent water layer provided by the
Global Surface Water Explorer (GSWE, https://global-surface-water.appspot.com/). The
flooded fraction was aggregated to district boundaries, and then the percentage flooded area
was calculated with respect to the district area.

c. Population

8
The gridded population data at 30-meter resolution was sourced from the Facebook Data for
Good platform (https://dataforgood.facebook.com), which provides high resolution population
density maps and demographic estimates. This data was re-gridded to the district level to
develop latest estimates of population. The impacts of flooding on a district having bigger
population will be more harmful compared to another district having smaller population.

3.1. District and State-Wise Floods

The total number of flood events per district and per state is presented in Figure 3. The district
of Thiruvananthapuram has experienced more than 231 occurrences of floods, or more than 4
events per year on an average. Further, 5 districts had more than 178 occurrences of floods, or
more than 3 events per year on an average. These are Thiruvananthapuram, Lakhimpur,
Dhemaji, Kamrup, and Nagaon. Also, 28 districts had more than 113 flooding events, or more
than 2 events per year. These are Thiruvananthapuram, Lakhimpur, Dhemaji, Kamrup, Nagaon,
Mumbai, Barpeta, Sonitpur, Nagpur, Darrang, Dibrugarh, Cachar, Wayanad, Goalpara, Baksa,
Pune, Jorhat, Morigaon, Nalbari, Golaghat, Sivasagar, Dhubri, Warangal Rural, Hamirpur,
Idukki, Alappuzha, Kozhikode, and Shimla. Furthermore, 101 districts had more than 56 flood
occurrences, which implies that parts of the district faced an average of one event per year. We
may categorize districts that experience at least one flood a year as flood prone, and districts
that face, on an average, more than three flood events per year to be severely flood prone. Figure
2(a) shows district wise occurrences of floods in India. It is noted that most flood prone districts
are in the Brahmaputra, Ganga basins, and some parts of Maharashtra and Kerala. Figure 2(b)
shows state-wise occurrences of flooding events in India [1967-2023] and the States in the
North-Eastern region where Brahmaputra and Barak rivers are flowing are severe flood prone.

9
Figure 3: (a) District wise occurrences of floods in India, and (b) state-wise occurrences of flooding events in India [1967-
2023]

The number of flooding events was also aggregated state-wise and presented in Figure
4 (a) for districts and Figure 4(b) for states. Expectedly, the state of Assam in North-East
India experiences the highest number of floods India, having faced more than 800 flood
events over 56 years. This implies that that every year, on an average, the state of Assam is
likely to face at least 14 flood events, highlighting the hardships that the people face there,
year-after-year. The other flood prone states in descending order of flood events include
Kerala, Karnataka, Uttar Pradesh, and Maharashtra.

10
 Figure 4: Top 30 districts in India with the highest number of floods, and (b) State-wise list of highest number of flooding
events [1967-2023]

3.2. India Flood Inventory – Death and Damages

The death and damage information in the India Flood Inventory can be considered the best
available lower bound of the actual losses as shown in Figure 5. One may note from this figure
is that due to improved flood management, human fatality is nearly stable or is coming down
in recent times despite increase in population in India. The number of deaths is around 1000
per year in recent times and efforts are underway to further reduce this number. Experience
from recent flood disasters reveals a few reasons behind large number of deaths are: dwellings
unsafe or building homes on unstable hillslopes, houses or too close to rivers or in flood plains,
landslides due to intense rainfalls, accentuating flood problem, and crossing rivers in spate
without adequate safety gears.

Figure 5: Evolution of deaths due to flooding events [1967-2023]

4. A District Flood Severity Index (DFSI) for India

Since a district is the functional administrative unit of India, decisions about disaster
management are generally taken at the district level. Decisions about relief and fund
disbursement require an accurate country-wide understanding of floods. Thus, we are
proposing a District Flood Severity Index (DFSI), with the objective of having a simple index
that can be maintained as a spreadsheet by the government. While more sophisticated
conceptions of flood severity are possible, the objective here is to develop an index that is
easy for the India Meteorological Department (IMD) to maintain in the long term. It is

11
expected that this index would be useful for planning, infrastructure construction, relief and
rescue organizations.
We define DFSI as
(1 + 𝐹𝑙𝑜𝑜𝑑 𝑒𝑣𝑒𝑛𝑡𝑠) × (1 + Mean Flood Duration) × (1 + 𝐹𝑎𝑡𝑎𝑙𝑖𝑡𝑦)#
DFSI = 𝑙𝑜𝑔!" ) L
× (1 + 𝐼𝑛𝑗𝑢𝑟𝑒𝑑) × (1 + Flooded Area) × Population)

(1)
where,
a) Flood Events = Number of flooding events in a district,
b) Mean Flood Duration = Mean of the flood duration of all flooding events in a district,
c) Fatality = Number of human fatalities in a district,
d) Injured = Number of humans injured in a district,
e) Flooded area = Percentage of area of a district that is historically flooded. Calculated as
mean of the flooded fraction from the 40 year reanalysis of the Indian Land Data
Assimilation System (ILDAS),
f) Population = district population.

Composite DFSI consists of six factors that can be classified in two groups of variables: a)
variables representing the occurrence of floods – number of flood events, mean flood duration,
flooded area, and population; b) variables representing the damage due to floods – human

12
fatality and the number of humans injured in a district. High weightage is given to fatality and
log transformation is used to scale the index in a range which is intuitively easy to visualize.

The average flooded area of a district as a percentage of total area is given in Figure 6. Some
districts in Assam, Bihar, Eastern UP, and isolated districts around the country emerge as
districts with highest percentages of their area affected by floods.

Figure 6: Average flooded area as percentage of a district’s total area calculated from the Indian Land Data Assimilation
System (ILDAS) reanalysis

Finally, using Eq. (1), we develop the District Flood Severity Index (DFSI), as given in the
Figure 7.

13
 Figure 7: District Flood Severity Index of India developed using the India Flood Inventory [1967-2023]

The top 30 districts of India with the highest DFSI are listed in Table 2. Patna (Bihar)
emerges as the most affected district in terms of flood severity.

Table 2: Top 30 districts with the highest District Flood Severity Index

DFSI District States DFSI

Rank
1 Patna Bihar 19.37
2 Murshidabad West Bengal 19.01
3 Thane Maharashtra 18.88
4 North 24 Parganas West Bengal 18.86
5 Guntur Andhra Pradesh 18.84
6 Nagpur Maharashtra 18.70
7 Gorakhpur Uttar Pradesh 18.40

14
 8 Ballia Uttar Pradesh 18.38
9 Purba Champaran Bihar 18.38
10 Purba Medinipur West Bengal 18.37
11 Muzaffarpur Bihar 18.37
12 Lakhimpur Assam 18.35
13 Kota Rajasthan 18.33
14 Aurangabad Bihar 18.33
15 Maldah West Bengal 18.28
16 Rajkot Gujarat 18.25
17 Prayagraj Uttar Pradesh 18.19
18 Aurangabad Bihar 18.18
19 Bahraich Uttar Pradesh 18.17
20 Cachar Assam 18.16
21 Ahmadabad Gujarat 18.13
22 Jalpaiguri West Bengal 18.11
23 Darjeeling West Bengal 18.09
24 Dibrugarh Assam 18.08
25 Azamgarh Uttar Pradesh 18.07
26 Chamoli Uttarakhand 18.06
27 Pashchim Champaran Bihar 18.02
28 Amravati Maharashtra 17.92
29 Medinipur West West Bengal 17.87
30 Samastipur Bihar 17.85

The Chamoli district is a hilly district in Uttarakhand. It does not face recurrent flood
problem but appears in this list because of a few isolated highly damaging flood events. Further,
among the list of 30 districts, 17 are in the Ganga basin and 3 in the Brahmaputra basin. Among
all the Indian river basins, human population is highest in the Ganga basin and high flood
proneness of this basin is worrisome for professionals and decision makers. Additionally,
Brahmaputra is a transboundary basin that affects four countries, and comprehensive flood
mitigation in Brahmaputra river basin would require national efforts and regional cooperation.
Because of high population pressure, some people live too close to the river which also provides
them their livelihood. These people are often reluctant to move to safer locations even during
floods and become highly vulnerable.

5. Conclusions and Future Work

Analysis of data compiled in this study has confirmed that floods are the worst natural
disaster that continues to hit India repeatedly. The flood problem is likely to become more
severe due to climate and land use/cover changes. The present work has created, for the first
time in India, a DFSI at a fine spatial scale by using data pertaining to six indicators
representing flood occurrences and damage caused. We have computed DFSI at the district

15
level which is an administrative unit for planning and development activities. Further, the data
used here was available at this scale and we have employed GIS and other modern tools to
process and store the data. We hope that this index would be useful for planning and decision
making. Based on time series of DFSI maps prepared for different years would reveal trends
in flooding and better targeted flood management.
To take this work forward, better data on more variables would be necessary. The
authors did not have access to data about economic damages due to floods. Such data are
collected for providing compensation to the victims of floods but comprehensive economic
loss data at finer resolutions is not easily accessible. The spatial data about floods needs to be
collected by laying down protocols. Use of remote sensing data along with ground truthing (to
verify values and to fill gaps) and storage of this data in modern geospatial databases would
make it easier to analyze the data.

Acknowledgements
This research was conducted in the HydroSense lab (https://hydrosense.iitd.ac.in/) of IIT
Delhi and the authors acknowledge the IIT Delhi High Performance Computing facility for
providing computational and storage resources. Dr. Manabendra Saharia gratefully
acknowledges financial support for this work through grants from ISRO Space Technology
Cell (STC0374/RP04139); Ministry of Earth Sciences Monsoon Mission III (RP04574);
Ministry of Earth Sciences (RP04741); and DST IC-IMPACTS (RP04558). Authors
gratefully acknowledge the Indian Meteorological Department (IMD) for providing access to
the datasets.

Compliance with Ethical Standards

The authors declare that they have no known competing financial interests or personal
relationships that could have appeared to influence the work reported in this paper.

Author Contributions
MS: Conceptualization, Writing, Editing
SJ: Conceptualization, Writing, Editing
VP: Analysis
HM: Data Curation
OPS: Original datasets

16
Data Availability
• The India Flood Inventory with Impacts (IFI-Impacts) database is published in
Zenodo: https://doi.org/10.5281/zenodo.11069835
• Analysis notebooks: https://github.com/msaharia/India-flood-inventory-impacts
References:
CWC (2013). Flood prone areas in the India. Central Water Commission, Govt. of India, New
Delhi.
Diakakis, M., Deligiannakis, G., Antoniadis, Z., Melaki, M., Katsetsiadou, N.K.,
Andreadakis, E., Spyrou, N.I., Gogou, M., 2020. Proposal of a flash flood impact
severity scale for the classification and mapping of flash flood impacts. J. Hydrol. 590,
125452. https://doi.org/10.1016/j.jhydrol.2020.125452
Gelaro, R., McCarty, W., Suárez, M.J., Todling, R., Molod, A., Takacs, L., Randles, C.,
Darmenov, A., Bosilovich, M.G., Reichle, R., Wargan, K., Coy, L., Cullather, R.,
Draper, C., Akella, S., Buchard, V., Conaty, A., da Silva, A., Gu, W., Kim, G.-K.,
Koster, R., Lucchesi, R., Merkova, D., Nielsen, J.E., Partyka, G., Pawson, S., Putman,
W., Rienecker, M., Schubert, S.D., Sienkiewicz, M., Zhao, B., 2017. The Modern-Era
Retrospective Analysis for Research and Applications, Version 2 (MERRA-2). J.
Clim. Volume 30, 5419–5454. https://doi.org/10.1175/JCLI-D-16-0758.1
Getirana, A.C.V., Boone, A., Yamazaki, D., Decharme, B., Papa, F., Mognard, N., 2012. The
Hydrological Modeling and Analysis Platform (HyMAP): Evaluation in the Amazon
Basin. J. Hydrometeorol. 13, 1641–1665. https://doi.org/10.1175/JHM-D-12-021.1
Ghosh, A., Dey, P., 2021. Flood Severity assessment of the coastal tract situated between
Muriganga and Saptamukhi estuaries of Sundarban delta of India using Frequency
Ratio (FR), Fuzzy Logic (FL), Logistic Regression (LR) and Random Forest (RF)
models. Reg. Stud. Mar. Sci. 42, 101624. https://doi.org/10.1016/j.rsma.2021.101624
GOI (1980). Report of Rashtriya Barh Ayog (National Commission on Floods). Ministry of
Energy and Irrigation, Government of India, New Delhi.
Kanth, A.K., Chitra, P., Sowmya, G.G., 2022. Deep learning-based assessment of flood
severity using social media streams. Stoch. Environ. Res. Risk Assess. 36, 473–493.
https://doi.org/10.1007/s00477-021-02161-3
Kumar, S.V., Peters-Lidard, C.D., Tian, Y., Houser, P.R., Geiger, J., Olden, S., Lighty, L.,
Eastman, J.L., Doty, B., Dirmeyer, P., Adams, J., Mitchell, K., Wood, E.F., Sheffield,
J., 2006. Land information system: An interoperable framework for high resolution
land surface modeling. Environ. Model. Softw. 21, 1402–1415.
https://doi.org/10.1016/j.envsoft.2005.07.004
Kuntla, S.K., Pandey, A. and Saharia, M. (2020). Developing a Flood Severity Index over
India. AGU Fall Meeting Abstracts.
Magotra, B., Prakash, V., Saharia, M., Getirana, A., Kumar, S., Pradhan, R., Dhanya, C.T.,
Rajagopalan, B., Singh, R.P., Pandey, A., Mohapatra, M., 2024. Towards an Indian
land data assimilation system (ILDAS): A coupled hydrologic-hydraulic system for
water balance assessments. J. Hydrol. 629, 130604.
https://doi.org/10.1016/j.jhydrol.2023.130604
Marchi, L., Borga, M., Preciso, E., Gaume, E., 2010. Characterisation of selected extreme
flash floods in Europe and implications for flood risk management. J. Hydrol., Flash
Floods: Observations and Analysis of Hydrometeorological Controls 394, 118–133.
https://doi.org/10.1016/j.jhydrol.2010.07.017

17
Niu, G.-Y., Yang, Z.-L., Mitchell, K.E., Chen, F., Ek, M.B., Barlage, M., Kumar, A.,
Manning, K., Niyogi, D., Rosero, E., Tewari, M., Xia, Y., 2011. The community Noah
land surface model with multiparameterization options (Noah-MP): 1. Model
description and evaluation with local-scale measurements. J. Geophys. Res.
Atmospheres 116. https://doi.org/10.1029/2010JD015139
Rashtriya Barh Ayog, 1980. Rashtriya Barh Ayog (National Commission on Floods) Report.
Government of India.
Saharia, M., Jain, A., Baishya, R.R., Haobam, S., Sreejith, O.P., Pai, D.S., Rafieeinasab, A.,
2021a. India flood inventory: creation of a multi-source national geospatial database to
facilitate comprehensive flood research. Nat. Hazards. https://doi.org/10.1007/s11069-
021-04698-6
Saharia, M., Kirstetter, P.-E., Vergara, H., Gourley, J.J., Emmanuel, I., Andrieu, H., 2021b.
On the Impact of Rainfall Spatial Variability, Geomorphology, and Climatology on
Flash Floods. Water Resour. Res. 57, e2020WR029124.
https://doi.org/10.1029/2020WR029124
Saharia, M., Kirstetter, P.-E., Vergara, H., Gourley, J.J., Hong, Y., Giroud, M., 2017.
Mapping Flash Flood Severity in the United States. J. Hydrometeorol. 18, 397–411.
https://doi.org/10.1175/JHM-D-16-0082.1
Schroeder, A.J., Gourley, J.J., Hardy, J., Henderson, J.J., Parhi, P., Rahmani, V., Reed, K.A.,
Schumacher, R.S., Smith, B.K., Taraldsen, M.J., 2016. The development of a flash
flood severity index. J. Hydrol., Flash floods, hydro-geomorphic response and risk
management 541, 523–532. https://doi.org/10.1016/j.jhydrol.2016.04.005

18