CLIMATE MODEL PRECIPITATION TREND ANALYSIS IN THE 20 TH CENTURY Karen Rivas Key Acknowledgements: Xubin Zeng & Koichi Sakaguchi UA/NASA Space Grant Symposium.

CLIMATE MODEL PRECIPITATION TREND ANALYSIS IN THE 20TH

CENTURY Karen Rivas

Key Acknowledgements: Xubin Zeng & Koichi SakaguchiUA/NASA Space Grant Symposium

April 21, 2012

1

Questions

Based on climate models how will global mean precipitation change?

How reliable are climate models in comparison with observational data in the past ~100 years?

2

Approach and Methods

Analyze various climate model precipitation trends

Analyze observational precipitation trends from two distinct set of data

Draw conclusions from comparison of model precipitation trends and observational trends

3

Climate Models

“Climate models use quantitative methods to simulate the interactions of the atmosphere, oceans, land surface, and ice.”

The use of climate models ranges from studying earth weather dynamics, to making predictions about future climate.

Model accounts for incoming and outgoing energy due to electromagnetic radiation

4

Data Analyzed

Models: NCAR-CCSM-4 21st Century Projections (future) NASA-GISS-E2-R 21st Century Projections

(future) NCAR-CCSM-4 20th Century Hindcasts (past) NASA-GISS-E2-R 20th Century Hindcasts (past)

As used in the coming IPCC Report

Observational Data: Global Land Precipitation,

Climatic Research Unit (CRU)

5

Analysis Techniques

Global mean precipitation is used to calculate:

change over ~100 year periods

respective change in 10-year windows

6

Blue: Global mean annual precipitation from GISS E2-R

Green: Linear fits for different 10-year periods

Results

7

Average annual precipitation rate

Change

Observational 1062.53 +35.01

GISS E2R 1158.85 - 6.78

CCSM4 1073.44 +13.55

Global Average Annual Precipitation (mm/year) and its Change over Historical Period 1900-1998

Results

8

Error GISS E2R CCSM4

Correlation (R) 0.46 0.10

R2 0.21 0.01

Root Mean Square Error (RMSE)(mm/year/decade)

49.60 51.57

Comparison between Models and CRU data for 10 year windows from Historical Period

Future Projections

9

Models 20th century 21st century

GISS E2R - 6.8 +25.6

CCSM3 +13.6 +34.8

Change of annual precipitation rate (mm/year) over the whole 20th and 21st

centuries

Annual precipitation (black) and10-year linear fits (colors)

Conclusion

Over the entire 20th century observational global mean precipitation has slightly increased

Over the entire 21st century both models project an increase in global mean precipitation

For the 10 year timescale, trend is highly variable and the model performance is poor for the 20th century.

10

Future Studies

Look at smaller spatial scales

Look at longer temporal scale

Look at more models

11

Thank You

Karen RivasDepartment of Chemical and Environmental Engineering

University of ArizonaApril 21, 2012

Tucson, Arizona

12