Practices of Downscaling Methods for Water Resources Management in Sri Lanka

Sewwandhi Chandrasekara

Sanjaya Rathnayake

Practices of Downscaling

Methods for Water Resources

Management in Sri Lanka

Introduction

• Sri Lanka owns different types of water

resources .Click>>

• Historically Sri Lanka had great hydraulic

civilization

• Presently water resources management is

essential to Sri Lanka

– 103 river basins (4560 km)

(Ministry of Forestry and Environment, 1999)

– More than 20 major

wetlands

– Minor and major irrigation

systems

– Groundwater resources

– Coastal and marine

resources. Click>>

• Sri Lanka

– 2400 m3 of available per

capita water resources

– 2000 mm of average

annual RF (Ariyabandu, 2008) click>>

Click>>

Historically

• Great ancient hydraulic civilization in Sri

Lanka

– Water used for

• Irrigation

• Domestic requirements

• Urban recreations

• Small scale industries

– However

• Water was managed properly

– Sustainably

– Participatory

But…..

• Temporal and spatial water scarcity

• Pollution

• Agrarian demand to sectoral demand. Click>>

– Industrial demand

– Environmental demand

– Electricity demand

– Recreational demand

Sustainable water resources management is

essential… Click>>

Click>>

Why prediction is important?

Threats for the water management

Statistical Downscaling…

• Downscaling for

– Rainfall OND

• Second inter-monsoon period

• 16% of quantity

• Important to land preparations for “Maha” season

(64% of paddy cultivation)

– Temperature OND

• Average 270C

Using SCM….

0.2 mm wetter than the normal

44% of success rate

Using SSE…

0.2 –0.4 mm wetter than the normal

44% of success rate

Using MRG….

0.2 –0.8 mm wetter than the normal

Poor success rate

Using SSE….

0.2 0C warmer than normal

Poor success rate

Using MRG…

0.2 0C warmer than normal

44%-55% success rate

Using SCM…

0.2 0C warmer than normal

55% success rate

Conclusions

• For OND season SSE & SCM can used to

predict rainfall at 44% success rate

• For OND season SCM can used to predict

temperature at 44% success rate

Unable to predict rainfall for the next season, because the stations what

we selected were failed at screening

Using Dynamic Downscaling

• Software : RegCM 4

• Predicted Region : Sri Lanka

• Latitude : 5 N – 10 N

• Longitude : 75 E – 85 E

• Duration : 1998 June

Westerly Wind

Southerly Wind

Geo Potential Height

Air Temperature

Relative Moisture

Water Vapor

Cloud Water

Surface Pressure

Sea Level Pressure

Total Precipitation

Total Soil Water in mm H2O

Accumulated Infiltration

Conclusion

• RegCM 4.0 is very usable to predict

South-west monsoon to Sri lanka

• Because past experiences showed the

same results observed from the RegCM

4.0 for the same monsoon season

Further Work

• Simulate long term predictions via

RegCM3

• Upload more applicable Sri Lankan station

precipitation data for CLIK on-line

application

Acknowledgement

• Dr. Lareef Zubair and staff at Foundation

of Environment and Climate Technologies,

Digana. Sri Lanka.

• Organizing Committee and Lecturers

APEC Climate Center. South Korea.

• Colleagues who participated with us.

Sewwandhi Chandrasekara

Sanjaya Ratnayake

(Sri Lanka)

APEC Climate Center

South Korea

Thanking you!