IUKWC Workshop November 2016: Developing Hydro-climatic Services for Water Security Session 4.1 Ashok Kumar Des

1. Operational Hydro-meteorological
Services
Dr. Ashok Kumar Das

2. HYDROMETEOROLOGY DIVISION New Delhi
LAND MARKS
 1890- Govt. of India passed the Rainfall Resolution of
India which made India Meteorological Department as
controlling authority for all rainfall measuring agencies.
 1949 - A special hydrometeorological unit was
established at Alipore, Kolkata to provide operational
met service to Damoder Valley Corporation.
 1963 – Storm Analysis Unit was established as per
Khosala Committee’ s recommendations
 1971 - The Hydromet Division is formally established
to cater increasing demands of value added information
to rainfall data as per users’ specific needs.

3. HYDROMET SERVICES
Hydromet
Activities
Hydro met Design
Rainfall
Monitoring
Hydro met
Forecasting (QPF)
for Flood
Forecasting

4. HYDROMET DESIGN

5. RAINFALL IN INDIA
 RAINFALL : 118.62 cm ( ANNUAL),
88.70 cm (SW Monsoon)
4.09 cm (Winter)
13.12 cm (Pre-Monsoon)
12.71 cm (Post-Monsoon)
(Data 1951-2000 used for Normal RF)
 HIGHLY VARIABLE IN SPACE
15% AREA OF INDIA > 150 CMS
64 % AREA OF INDIA 75 – 150 CMS
15% AREA OF INDIA 40 – 75 CMS
6% AREA OF INDIA < 40 CMS
 HIGHLY VARIABLE IN TIME
75 % OF ANNUAL RAINFALL OCCURES
DURING 4 MONTHS.
 HENCE STORAGE STRUCTURES ARE
NEEDED TO MANAGE BALANCE 8
MONTHS.

6. Hence we have to develop infra structure for full
exploration of available potential.
Water Availability in India
Area of the country 32,87,263 Km2
Total utilizable water resources 1123 BCM
Surface water potential 690 BCM per year
Ground water potential 433 BCM
Expected water demand 1050 BCM (by YEAR 2025 )

7.  At the end of 2011, there are more than 4845 large* dams,
and 347 under construction to meet the demand of power
generation and Flood Control.
 The total storage capacity of these dams, is about 283
billion M3.
 Because of the huge investment, safety of these
structures, Requirement of Water Supply, Hydro-electrical
power need to be insured.
 The main cause of dam failures is over toppling by
excess inflow.
 Hence, spillway should have adequate out flow capacity
for Dam safety
(* height ≥ 15 M or storage Capacity > 1 Million M3)
(Source: National Register of large dams, compiled by CWC, MoWR up
to Feb. 2012)
Requirement of Dam safety

8. FEW EXAMPLES OF DAM FAILURE DUE
TO INADEQUATE SPILLWAY CAPACITY
KADDAM DAM 1958
PANSHET & KHADAKWASLA DAM 1961
MACHHU PROJECT – II 1979
KOSI BARRAGE 2008
DAM FAILURE

9. Design Floods are estimated from Design
storm studies.
In order to cater this need, Hydromet
Division of IMD undertakes the Design
Storm studies comprising of
“Estimation of Probable Maximum
Precipitation (PMP)”
with the help of isohyetal analysis of
rainfall
DESIGN STORM STUDIES

10. The Probable Maximum
Precipitation is obtained
from
- Isohyetal Analysis of
historic rain storms.
- Transposition of
artificially selected
rainstorm to the project
basin.
- Further maximizing the
rainstorm to extreme
moisture condition etc. to
arrive at PMP
DESIGN STORM STUDIES

11. In case of minor structures like
bridges, culverts, planning of water
supply system & drainage systems,
railway and road bridge
construction etc.,
IMD provides
 Estimates of short duration rainfall
return period values
 Isopluvial maps
 Intensity Duration Frequency (IDF)
curve
For Govt. agencies, these studies
are being carried out free of cost
and for private / profit earning
agencies on payment basis.
DESIGN STORM STUDIES

12. RAINFALL MONITORING

13. Information Disseminated
to USERS
Prime Minister House
Cabinet Secretary
Agriculture Ministry
State Agencies
Finance Ministry
Foreign Embassies
Insurance Companies
CWC, NCMRWF & NRSA
Rain Recording
Station
Station R/F transmitted to R.M.C. / M. C.
Regional /
Met.
Centre
Districtwise
R/F statistics
is prepared
Districtwise
R/F is
transmitted
to
Hydromet
Division,
New Delhi.
HYDROMET
Prepares R/F statistics
for the Subdivisions,
States, Regions and
the Country
DATA PROCESSING FOR RAINFALL SUMMARY REPORTS
 In 1989 the Districtwise Rainfall Monitoring Scheme (DRMS) got approved
for compiling rainfall summary for all the seasons and for Districts / met sub
Divisions and Country as whole.
IMD are compiling District / met sub divisional and Country’s realtime rainfall
summary for appraisal of Rainfall situation on daily/ weekly/ monthly/ seasonal/
annual basis.
 About 3500 stations
data are being used
for preparation of real
time rainfall summary.
The rainfall
summary is thus
updated on monthly
basis after
incorporating late
data at RMC /MC
level.

14. Scheme of computations.
 The estimate of District rainfall is made by Arithmetic Mean Method
RDist = ∑ Ri / N
 It was then thought of estimating sub divisional average rainfall
from Arithmetic Method considering all the stations’ rainfall in the
sub Division in same way like district rainfall.
But it was noticed that there is bias towards the area where
density of rain gauges is higher than other parts.
 Hence met sub Divisional rainfall estimates are made taking Area
Weighted Average Rainfall of Districts of that met sub Division
RsD = ∑ RDist*ADist / AsD
 The rainfall summary for country as whole is compiled as area
weighted average rainfall of met sub Divisions in similar way.
Rcountry = ∑ RsD*AsD / Acountry

15.  For given week (period), the summary for dist., Met. Sub Division and
Country comprises of
- Actual Rainfall Estimate
- Normal Rainfall for given period (based on 1951 -2000 rainfall data)
- Departure from Normal = (Actual R/F – Normal R/F) / Normal R/F
(expressed in terms of %)
- Category
Excess – When Departure > 19%
Normal -19% ≤ Departure ≤ 19%
Deficient -59% ≤ Departure < -19%
Scanty -99% ≤ Departure < -59%
No Rain Departure = -100%
Scheme of computations.

16. Sub Division wise and State wise Maps

18. DISTRICTWISE RAINFALL MAP

19. DISTRICTWISE RAINFALL MAP

20. DISTRICTWISE RAINFALL MAP

21. DISTRICTWISE RAINFALL DISTRIBUTION

22. CRIS (CUSTOMISED RAINFALL INFORMATION SYSTEM)
THE SYSTEM CRIS WHICH WAS MADE OPERATIONAL ON 15TH JANUARY, 2015 ON IMD WEBSITE
PROVIDES REAL-TIME RAINFALL INFORMATION BY MEANS OF GIS BASED RAINFALL PRODUCTS
FOR ANY PLACE IN INDIA.
Rainfall
Maps
Rainfall
Graphs
Rainfall
Statistics
Rainfall
Departures
Spatial
Analysis
CRIS PRODUCTS

23. Computation of River basin wise rainfall
statistics(actual, normal & % dep.) which is
useful for flood forecasting and Water
Resources management
New Products on CRIS
(River Basin wise information)
RIVER BASIN-WISE SPATIAL
ANALYSIS OF OBSERVED RAINFALL
IS USEFUL FOR FLOOD
FORECASTING AND WATER
RESOURCE MANAGEMENT

24. Real time Hourly cumulative (one day) AWS/ARG
rainfall data on Pan India. Useful for real time
rainfall information, Urban flooding etc.
REAL TIME STATIONWISE RAINFALL DATA ON CRIS

25. Products on CRIS
Cumulative Rainfall distribution. Useful for
real time rainfall information
State-wise Rainfall map
District wise rainfall map for a State
(Available for all states)
THE DISTRICT-WISE
RAINFALL STATISTICS IS
HELPFUL TO FARMERS FOR
THEIR AGRICULTURAL
ACTIVITIES.

26. Products on CRIS
WEEK-BY-WEEK DEPARTURES OF
RAINFALL FROM NORMAL HELP
VARIOUS STAKE HOLDERS (LIKE GOVT.
AGENCIES, FLOOD AND DROUGHT
MITIGATING AGENCIES, INSURANCE
AGENCIES ETC.) FOR ESTIMATING
DROUGHT AND FLOOD CONDITIONS
IN THE COUNTRY.
SEASONAL & ANNUAL NORMAL RAINFALL (1951- 2000 )
MAPS OF STATE, MET SUB DIVISION AND RIVER BASINS

28. CUSTOMIZED RAINFALL INFORMATION SYSTEM
(CRIS)

29. Hydro met Forecasting (QPF) for Flood Forecasting

30. NATURE OF FLOODS IN INDIA
Maximum floods Occur in S-W MONSOON season.
Flood is an ANNUAL FEATURE of major rivers.
According to the estimate of the National
Commission on Flood, the area prone to floods in
the country was of the order of 40 million
hectares, out of which it is considered that 80%,
i.e., 32 million hectares could be provided with
reasonable degree of protection.
On an average75000 SQ. KMS of area & 33 million
population faces FLOOD WRATH every year.

31. .........
.
.
.
.
PAKISTAN
88o 72o
76o
80o 84o 88o 92o 96o
36o
32o
28o
24o
20o
16o
12o
8o
92o88o84o80o76o72o
8o
12o0
16o0
20o
24o
28o
32o
36o
HIMACHAL
PRADESH
UTTAR
PRADESH
HARYANA
RAJASTHAN
GUJRAT
MADHYA PRADESH
MAHARASHTRA
BIHAR
ORISSA
TRIPURA
MIZORAM
MANIPUR
NAGALANDASSAM
CHINA
TIBET
BHUTAN
Narmada
Saraswati
Mahi
ANDHRA
PRADESH
TAMIL NADU
KARNATAKA
KERALALAKSHA DWEEP
(INDIA)
INDIAN
SRI
LANKA OCEAN
MYANMAR
Tasango
BAY
OF
BENGAL
ANDAMAN AND NICOBAR ISLANDS
(INDIA)
GOA
ARABIAN
SEA
SRINAGAR
NEW DELHI
JAIPUR
GANDHINAGAR
BHOPAL
MUMBAI
HYDERABAD
PANAJI
CHENNAIBANGALORE
KAVARATI
THIRUVANANTHAPURAM
PORTBLAIR
BHUBANESHWAR
AIZAWALAGARTALA
IMPHAL
KOHIMA
SHILLONG
DISPUR
ITANAGAR
GANGTOK
MEGHALAYA
KOLKATA
ARUNACHAL
PRADESH
SIKKIM
BANGLADESH
FLOOD PRONE AREA
Kilometres100 50 0 100 200 300 400 Kilometres
Miles100 50 0 100 200 300 Miles
SCALE
JAMMU & KASHMIR
SHIMLA
CHANDIGARH
WEST
BENGAL
PATNA
NEPAL
LUCKNOW
PUNJA
B
Flood Prone
areas in
India
~ 40 million Ha
(RBA,1980)
 Major Flood
Prone States:
Assam,
Bihar,West
Bengal, UP,
 Odisha &
Andhra Pradesh
 Major Flood
Prone Basins:
Ganga,
Brahmaputra &
Mahanadi

32. In past 50 years
India had lost
assets of more
than RS 650000
million (@ US$
14800 million) and
loss of more than
75000 human
lives due to
floods alone.
FLOOD DAMAGES

34. EXPERTS OPINION
Though total immunity from floods is not feasible, the
losses particularly of lives and movable property can be
minimized by
 Structural Measures
 Non Structural Measures
Structural Non Structural
Construction of Dams,
Embankments etc
Flood Warnings
Flood plain zoning
etc
Costly Relatively less costly
Saves both movable and
immovable property
Saves only movable property.

35. CAUSES OF FLOOD
 HEAVY RAINS IN SHORT PERIOD
 LANDSLIDES AND AVALANCHES
 CHANGE IN RIVER COURSE
 FAILURE OF DAMS AND OTHER HYDRAULIC
STRUCTURES
 ENCROACHMENT IN FLOOD PLAIN AREAS
 CYCLONES,THUNDERSTORMS, CLOUD BURSTS
 GLACIAL OUT BURST
However majority of floods occur due to
Heavy rains in short period. Hence accurate
QPF is a prime factor of reasonable Flood
forecasting and warning.

36. BREACH
IN
KOSI
BARRAGE
IN
AUGUST 2008
CAUSES OF FLOOD

37. Flood Forecasting
MINISTERIAL COMMITTEE ON FLOODS &FLOOD RELIEF-1972 RECOMMENDED
THAT
 Flood Forecasting Centres should be established in all flood prone basins.
 Flood Meteorological Offices should be established by India Meteorological
Department to work in close co-operation with Flood Forecasting Centres
FLOOD
METEOROLOGICAL
OFFICE (FMO)
INDIA
METEOROLOGICAL
DEPARTMENT
COORDINATION
CENTARL FLOOD
FORECASTING
DIVISION (CFFD)
CENTRAL WATER
COMMISSION
QUANTITATIVE
PRECIPITATION
FORECAST
FLOOD
FORECAST
FLOOD FORECASTINGIn India Flood Forecasting
activity is the joint
responsibility of
India Meteorological
Department
 Central Water
Commission.
IMD provides Hydromet Input
through their Flood Forecasting
Offices(FMOs).
In hydromet input, Quantitative
Precipitation Forecast (QPF) is
the main input which is used in
rainfall runoff models by CWC.

39. Sl. No. FMOs Date of Estt. River Basins
1 Agra 29-Mar-1985 Banganga, Chambal, Betwa, Ken, Yamuna
2 Ahmedabad Jan-1974 Narmada, Tapi, Daman-Ganga, Sabarmati, Banas, Mahi
3 Asansol Mar-1980 Ajoy, Mayurakshi, Kangsabati
4 Bhubaneswar 11-Jan-1974
Subarnarekha, Baitarni, Burhabalang, Vamsadhara,
Brahmani, Mahanadi, Rushikulya
5 Guwahati Aug-1975
Brahmaputra, Barak, Dehung, Lohit, Buridihing, Subansiri,
Dhansiri (N,S), Jiabharali, Kapili, Manas/ Beki, Sankosh
6 Hyderabad Apr-1977 Godavari, Krishna
7 Jalpaiguri 24-Jul-1974 Teesta, Jaldhaka, Raidak
8 Lucknow Sep-1974
Ghaghra, Rapti, Ramganga, Gomti, Sai, Sahibi, Chhatang,
Bhagirathi, Alaknanda, Ganga
9 New Delhi 15-Jun-1974 Yamuna, Sahibi
10 Patna Dec-1973
Kosi, Mahananda, Adhwara, Bagmati, Gandak, Punpun,
Sone, Kanhar, North Koel
11 MC Srinagar Jun-2015 Jhelum, Dah, Nimmo, Khalsi
12 RMC Chennai Jun-2016 Pennar, Lower Cauvery, Vaigai, Vellar
13 MC Bengaluru Jun-2016 Krishna, Cauvery, Tungabhadra
14 DVC 1949 Barakar, Damodar, Lower Valley

41. SYNOPTIC
ANALYSIS
SATELLITE
PRODUCTS
RADAR
PRODUCTS
ANALYSIS
QPF
Inputs for issuing of QPF
SYNOPTIC
ANALOGUE
NWP
MODEL

42. DURING FLOOD SEASON, FMOs send QPF Bulletin (9.30AM) and HYDROMET
BULLETINS (12.30PM) to CFFD which includes following :
 CATEGORICAL SUB BASIN WISE QUANTITATIVE PRECIPITATION FORECAST (QPF)
FOR THE NEXT THREE DAYS AND OUTLOOK FOR FURTHER FOUR DAYS FOR THE
CONCERNED CATCHMENTS/SUB-CATCHMENTS
 PREVAILING SYNOPTIC SITUATIONS
 SUB BASIN WISE AREAL RAINFALL OCCURRED DURING PAST 24-HOURS
 STATIONWISE SIGNIFICANT RAINFALL DURING THE PAST 24-HOURS
 HEAVY RAINFALL WARNING FOR THE NEXT 72-HOURS
HYDROMET BULLETIN
Other than flood season, QPF is to be issued to CFFD under following
eventualities:
During the period of ‘Flood Alert’ for specified basins as notified by
CFFDs.
If heavy rainfall is expected during non-flood season or in the event of
formation of Cyclone, concerned FMO will issue QPF/HM Bulletins.

43. FMOs ALSO PERFORM FOLLOWING FUNCTIONS DURING NON
FLOOD SEASON :
 PREPARATION OF SYNOPTIC ANALOGUE MODEL FOR QPF
 PREPARATION OF HYDROMETEOROLOGY OF RIVER BASINS
 FMO STATIONS ARE BEING INSPECTED BY THE RESPECTIVE
FMOs EVERY YEAR FOR ACCURATE AND EFFICIENT
RECEPTION OF RAINFALL DATA ON REAL TIME BASIS
FLOOD METEOROLOGICAL OFFICES

44. Flood Forecasting: Basic Structure
.
Hydrological ModelHydrological
Observations
Reception of
Rainfall data from
RCs/MCs/FMOs
Weather
Forecasts(QPF)
SYNOPTIC CHART, SATELLITE& RADAR
PRODUCTS, SYNOPTIC ANALOGUE &
Weather Forecast Models
FMO
CFFD
Flood
Warnings

45. OUTPUT PRODUCTS
OF WRF
OUTPUT
PRODUCTS
OF MME
OUTPUT
PRODUCTS
OF GFS
Following sub basin wise model rainfall
forecast products are uploaded in the
IMD website;
WRF(ARW) (9km x 9km) (00utc & 12
utc),
Multi-Model Ensemble (27km x 27km)
(00utc)
GFS (22kmx22km) (00utc)
SUB BASIN WISE OPERATIONAL WRF, MME AND GFS

46. QUANTITATIVE PRECIPITATION FORECAST
The performance of
operational QPF and
dynamical NWP model
forecast are verified
during the monsoon
season by computing
different skill score
tests viz. Percentage
Correct Forecast(PC),
Heidke Skill Score
(HSS) and Critical
Success Index (CSI)
from 6X6 Contingency
table.

47. QUANTITATIVE PRECIPITATION FORECAST
 The POD, FAR, MR, C-NON,
CSI, BIAS, PC, TSS and HSS
for each category are
computed by reducing the 6X6
contingency table into 2X2
contingency table for
occurrence / non occurrence
(YES/NO).
 The computing procedure for
various skill scores tests are
as follows over sub basins.

48. DROUGHT MONITORING
On the recommendations of the Planning commission,
Drought Research Unit started functioning at Pune in
1967 in the office of the Additional Director General of
Meteorology (Research). The salient activities of this
Unit are as under:-
i. Defining meteorological Drought and its intensity.
ii. Delineation and identification of Drought Prone areas
of the country
iii. Study of past droughts and
iv. Monitoring Agricultural drought conditions during
Southwest and Northeast monsoons
v. Issuing weekly aridity outlook and
vi. Research on different aspects of drought.