satellite products for weather analysis and nowcasting in...

Satellite Products for Weather Analysis and Nowcasting in

Africa Estelle de Coning

Chief Scientist: Nowcasting and very short range forecastingSouth African Weather Service

Contributions by Eugene Poolman

1

Content

• Introduction• SWFDP• Measuring precipitation in DC/LDC• SAFFG and SARFFG• Summary and conclusion

2

1. Introduction• Strength of nowcasting - location-specific forecasts of the

initiation, growth, movement and dissipation of weather phenomenon

• Ideal world - radar systems • Reality, however, is that many developing countries (DC) and

even more so Least Developed Countries (LDC) do not have operational radar systems at all, and the countries which are fortunate enough to have radar systems, are struggling to maintain and sustain these powerful data sources.

• The Global Humanitarian Forum states: – “Developing countries, which are most likely to suffer the brunt of climate

change impacts, have the least number of ground-level weather data observation systems, the critical basis for efficient delivery of weather information.

– Despite covering a fifth of the world's total land area, Africa has the least developed land-based weather observation system of all continents, and one that is in a deteriorating state.

– Many existing weather stations do not operate properly, or do not operate at all.

– WMO estimates that in an ideal scenario, 10 000 weather stations should be operating in Africa. Currently, there are only around 744 stations operational, less than a quarter of which provide observations that meet WMOrequirements for standard and frequency of data.”

Global Humanitarian Forum (2008). Weather information for all initiatives 2008-12 available at http://info.publicintelligence.net/WIFA_Project_Outline_Executive_Summary.pdf.

More than 152 DC world wide in 2008

LDC – 33 in Africa, 14 in Asia Pacific (2011)

Radar networks in DCs• China, Romania, Brazil, Vietnam, Cuba, Ukraine, UAE, Saudia Arabia,

Thailand, Turkey, Philippines, Mali, Kenya, Rwanda (is purchasing a new one), Burkina Faso, Mozambique, Botswana, South Africa (not all operational)

• Cuba has several very old Russian radars that are kept functional through the efforts of one dedicated radar engineer and these limited radars are very important for their basic operational detection of storms.

• The World Bank, UN, USAID, national science foundations and national weather services assist in financial support to extend radar coverage.

• “The East African nations recognize the power of weather radars and have plans to incorporate them into their weather observation networks around Lake Victoria.

• In Kenya, there are radars in place in Nairobi and Mombasa. While these two radars are no longer operational, the Kenya Meteorological Department is hopeful to bring them back online in the near future.

• With regards to the Lake Victoria region, the Kenyans have plans to place a S-band scanning weather radar at the Eldoret International Airport, located approximately 100 km northeast of the lake.

• The Tanzanian Meteorological Agency is currently running one scanning weather radar in Dar es Salaam, and is in the procurement stage of putting a weather radar in Mwanza, located directly on the south-central shore of Lake Victoria. They have an excellent site already selected which would provide coverage over much of Lake Victoria.”

(Atmospheric Observations Feasibility Study report by Burleyson, Yuter and Rose, 2011)

WMO Nowcasting symposium (Whistler, 2009)

• Challenges surrounding the support of nowcasting efforts in developing countries:– observational data availability, – IT infrastructure and maintenance, – tools for processing and visualization, – training of operational staff in nowcasting concepts, – end-to-end product dissemination, – building and maintaining regional radar networks, – access to satellite data and local/regional Numerical Weather

Prediction (NWP) model output.

Practical advice1. Plan purchasing of expensive system in order to be sustainable2. Plan the dissemination of datasets via the internet,

GEONETCAST etc.3. Start by getting access to satellite data…”it is an efficient poor

man’s nowcasting system”.4. Plan how to develop the expertise and technical support staff in

your service necessary to keep these systems operational.5. Develop and document a sustainable plan for incorporating very

short term forecasting in the operational setting. 6. Provide and document end-to-end training taking into account

end-user requirements and needs.7. Regional/international sustainable cooperation is crucial.

2. Severe Weather Forecasting Demonstration Projects (WMO CBS SWFDP)

• The SWFDPs aims:– to improve severe weather forecasting in DC and LDC

by providing access to current forecasting information (e.g. NWP and EPS).

– Improve lead time of warnings– Improve interaction between NMHS and media as well

as disaster managers and civil protection authorities– Identify areas to improve upon through regular

feedback


• Current status of SWFDPs around the world –1. Southern Africa

• South Eastern Africa 2006-2007 (5 countries involved)• Southern Africa expansion 2008-2011 (16 countries involved)• RSMC established in Pretoria, South Africa

2. South Pacific islands• Pilot project 2009-2010• Full demonstration 2011• RSMC established in Wellington

3. Eastern Africa initiated4. Under development –Southeastern Asia and Bay of Bengal

2. Southern Africa SWFDP• Feedback from the participating NMHSs in the Southern African SWFDP:

– improved warning services in many countries and – contributed to the improvement of relations between NMHSs and disaster

management authorities. • The SWFDP has thus contributed significantly to and end-to-end process

of warning dissemination.


• Fourth meeting of the CBS-SWFDP Steering Group in Geneva, February 2012:– Challenge for the SWFDP: “the need for very short-range forecasting

tools, to address especially the rapid onset of localized severe thunderstorms which can produce heavy precipitation and strong wind, given the absence of adequate real-time observational networks, especially weather radar coverage.”

– The usefulness of EUMETSAT satellite based instability products, such as the Global Instability Index, for nowcasting purposes was recognized

– Also agreed that real time satellite rainfall estimates have proven particularly useful in regions where rain gauges and radar coverage is sparse.

• Surface-based precipitation measurement systems in DC and LDC still needed to accurately measure and monitor precipitation amounts on the ground, to be incorporated into hydrological runoff models as well as aid in validation of other (satellite and model based) methodologies.

3. Measurement of precipitation

• Rain gauges • Radars • Satellite based estimates of rainfall - not as accurate

as gauges or radar, its major advantage is the high temporal resolution and coverage, even over oceans, in mountainous regions and sparsely populated areas where rainfall is not measured. IDEAL for DC and LDC!

15

The Hydroestimator• NOAA/NESDIS developed an automated SPE algorithm

for high-intensity rainfall called the Autoestimator (AE). • Another version of the AE, called the Hydroestimator

(HE) has been developed which can be used outside of regions of radar coverage without compromising too much accuracy.

16

Hydroestimator – operationally available every 15 minutes

17

HE Tracking• HE tracking done of the 10mm (~40 dBz

core)• Tracks storm direction for 30min and 60min

forecasts. • Growth and decay accounted for

18

HE accumulations

• 1 hour, 3 hour, 6 hours, 24 hours• 10 days• 30 days

19

4. South Africa Flash Flood Guidance (SAFFG)

• Hydro-meteorological modelling system combining :– meteorological information, such as quantitative rainfall estimation from weather

radars, satellite and rain gauges, – with hydrological modelling of the soil moisture conditions

• Result: flash flood potential in 5366 small river basins (on average 50 km2) in five flash flood prone regions over South Africa.

• The SAFFG uses the quantitative rainfall estimates of the previous hours from:– radar, – satellite and – rain gauges

• Calculates the amount of rain needed over the basin that will lead to bankfullat the outlet of the river, i.e. start of flooding on an hourly basis.

• The Flash Flood Guidance System (FFGS) is the intellectual property of the Hydrologic Research Center (HRC), a non-profit public-benefit corporation based in San Diego, USA. SARFFG was developed and implemented by HRC.

21

Flash Flood Guidance Process

22

Calculate from satellite & rain gauge info the averaged rainfall over small basins

Hydrological models determine likely soil moisture and rainfall runoff for small basins

22

This cycle is repeated every hour to update FFG using latest rainfall information

Determines potential for flash floods as guidance to forecasters

SAFFG

23

Towards improving satellite QPE for SAFFG

• Use 2 years of HE and rain gauge data – 2008 and 2009

• Determine the HE bias • Use UM stratiform data for the coastlines of

SA where cold fronts/ridging high pressure systems cause stratiform rainfall

• Determine model bias

24

Blended precipitation product • To account for convective as well as stratiform

rainfall processes….• Bias corrected HE and the bias corrected UM

statiform rainfall field are combined in to one new product (called COMB)

• COMB is dependent on the month of the year for different bias corrections

• COMB is the maximum of bias corrected HE and bias corrected UMS at a given pixel to ensure extreme events are not missed.

De Coning, E and E Poolman. April 2011 South African Weather Service operational satellite based precipitation estimation technique: applications and improvements. Hydrol. Earth Syst. Sci., 15, 1131-1145,

www.hydrol-earth-syst-sci.net/15/1131/2011/doi:10.5194/hess-15-1131-2011

25

Case: 18 June 2008

Raingauges

HE

UMS

Combination

26

Statistical evaluation for 18 Jun 2008

RED = CombinationBlue = HE

27

Statistical evaluation for 28 Jan 2008

28

RED = CombinationBlue = HE

Annual statistics 2008/9

29

0

2

4

6

8

10

12

14

Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec

Mean Absolute Error 2008

HE

Comb

0

1

2

3

4

5

6

7


Mean Absolute Error2009

HE

Comb

30

-0.05

0

0.05

0.1

0.15

0.2

0.25

0.3


Hanssen Kuiper score2008

HE

Comb

0

0.05

0.1

0.15

0.2

0.25

0.3


Hanssen Kuiper score2009

HE

Comb

Annual Statistics 2008/9

Published in April 2011: Hydrological and Earth System Sciences (HESS)

SAFFG for the rest of southern Africa (SARFFG)

• WMO - similar flash flood guidance system (called the SADC SARFFG) over seven southern African countries

• The SADC SARFFG system covers the rest of South Africa and six other countries where there are no radar coverage at the coarser 200km2 resolution.

• SARFFG depends primarily on satellite QPE as precipitation input for modelling soil moisture and flash flood guidance over large parts of southern Africa.

31

Further improvements

• Used 2 years for climatology for the bias correction, now adding another 2 years

• Will also improve on bias correction to be more area specific

• Investigating multi channel MSG QPE

33

Summary • Satellite based QPE – great spatial coverage• Hydroestimator – available every 15 minutes, as well as

accumulation and tracking products• SAFFG and SARFFG – guidance to forecasters for flash flood

forecasting fed by HE, radar rainfall (SA) and gauges• Blended precipitation product evaluates better than HE alone

and if this new field is used as input to the SAFFG, a better outcome is expected where/when stratiform plays a role

• Accurate satellite rainfall estimation is crucial for good guidance in data sparse areas over Africa

34

Summary• In the absence of or with limited radar systems and extensive

observation networks NWP and satellite based products can go a long way to help forecasters in DC and LDC to do nowcasts

• Systems such as SWFDP and SARFFG are web based and thus easy to access by forecasters in NMHSs to use the information received from the global and regional centers to issue nowcast information

• The importance of developing an excellent collaboration between weather forecasters and hydrologists, and between weather forecasters and disaster managers, in each country for the successful implementation of the SWFDP and SARFFG is and will remain, of course, fundamental.

• Ground based data sources (such as gauge networks and radar systems) – remain crucial/ideal not only for observation but for verification of effectiveness of other data sources.

satellite products for weather analysis and nowcasting in...

Documents