Regional Climate Modeling in the Source Region of Yellow River with complex topography using the RegCM3: Model validation

1
Regional Climate Modeling in the Source Region of
Yellow River with complex topography using the
RegCM3 Model validation

• Pinhong Hui, Jianping Tang
• School of Atmospheric Sciences
• Nanjing University, China

August 27-30, 2013 Katmandu, Nepal
2
Content

• Introduction
• Experiment design
• Observed data
• Simulation result
• Climatology
• Variability
• PDFs and quantiles
• Extreme indices
• Summary

3
Introduction

• The Tibet Plateau
• The highest plateau all over the world
• complex topography and fragile ecosystem
• one of the most sensitive areas to climate change
• The Source Region of Yellow River
• located in the Tibet Plateau
• climatology may have dramatic impact on hydrology
  and ecosystem over the whole Yellow River Basin
• Spatial distribution of precipitation and
  temperature
• displays a strong relationship with the
  topography in scale under 10km
• However, most of this region lacks meteorological
  observations
• Use of regional climate models(RCMs)
• necessary for reproducing the main climatic
  features in complex terrain

4
Introduction

• The regional climate models has been successfully
  applied in many regional climate studies around
  the world
• Heikkila et al. (2011) did dynamical downscaling
  of the ERA-40 reanalysis with the WRFV
• Afiesimama et al. (2006) use the RegCM3 to study
  the West African monsoon
• Dimri and Ganju (2007) simulated wintertime
  Seasonal Scale over Western Himalaya Using RegCM3
• Park et al. (2008) Characteristics of an
  East-Asian summer monsoon climatology simulated
  by the RegCM3
• Caldwell et al. (2009) Evaluation of a WRF
  dynamical downscaling simulation over California
• ...
• There is little research work on regional climate
  modeling in the Source Region of Yellow River
  with high resolution using the RegCM3 model

5
Experiment design
Model Configuration Model Configuration
Model prototype RegCM3
Governing equations Hydrostatic
Grids and resolution 11078, 45km 15km
Vertical layers (top) 18 sigma layers (50hPa)
Cumulus convection Grell
PBL Holtslag
Land Surface BATS
Initial and boundary conditions ERA-interim reanalysis
Simulation period Simulation period
1989.1.1-2009.12.31 1989.1.1-2009.12.31
6
Experiment design

• First figure the larger domain with 45km
  resolution covering the whole China with a 15km
  nest covering the Source Region of Yellow River
• Second figure
• shaded color terrain height in the nest
• large red rectangle analysis domain(92-106E,
  29-39N)
• black contour line location of the Source Region
  of Yellow River
• small red circles surface observation stations

7
Observed data

• Daily surface observations from the China
  Meteorological Administration (CMA)
• Precipitation
• surface air temperature
• daily maximum and minimum surface air temperature
• Consists of 756 meteorological stations, covering
  the whole country and provides the best data
  available for China
• 116 stations included in our analysis domain
• Interpolated the model results onto the station
  locations and evaluated the quality of the
  simulations

8
Simulation resultClimatology
9
precipitation bias
Statistical index Statistical index Whole region Whole region Source Region Source Region
Statistical index Statistical index Summer Winter Summer Winter
45km BIAS() 19.770 268.322 14.760 122.096
45km Spatial R 0.865 0.768 0.786 0.842
45km RMSE (mm/day) 3.412 3.999 1.869 2.005
15km BIAS() 13.368 227.880 12.055 80.987
15km Spatial R 0.883 0.792 0.873 0.880
15km RMSE (mm/day) 2.882 3.546 1.753 1.887

• high-resolution, remarkable improvement
• 15km simulation
• bias and RMSE, much smaller
• spatial correlation coefficient, much higher

• Overestimation , especially in winter
• Source Region of Yellow River, better simulated
• Largest bias, Qaidam Basin
• Underestimation, Tanggula Mountain, Sichuan Basin

10
surface air temperature bias
Statistical index Statistical index Whole region Whole region Source Region Source Region
Statistical index Statistical index Summer Winter Summer Winter
45km BIAS(?) -3.399 -3.961 -1.821 -1.713
45km Spatial R 0.703 0.544 0.687 0.313
45km RMSE(?) 0.677 0.386 0.593 0.202
15km BIAS(?) -2.867 -3.506 -1.704 -1.594
15km Spatial R 0.786 0.594 0.899 0.787
15km RMSE(?) 0.512 0.328 0.528 0.138

• 15km simulation outperforms the 45km simulation,
  especially in the Source Region of Yellow River
• higher spatial correlation coefficient
• lower bias and RMSE

• cold bias
• Maximum bias, surroundings of Tanggula Mountain
• locations of cold bias are in good agreement with
  the wet bias regions

11
Precipitation and surface air temperature at
different surface elevations
12

• Simulation resultVariability

13

• Inter annual variability of precipitation and
  surface air temperature averaged over the whole
  analysis domain

14
Taylor Diagram of interannual variability in the
12 surface stations in the Source Region of
Yellow River
15

• Annual cycle of precipitation and surface air
  temperature averaged over the whole analysis
  domain

16
Simulation result PDFs and quantiles
17

• PDFs of daily mean precipitation over the whole
  analysis region and the Source Region of Yellow
  River

18

• Quantiles (0.025, 0.1, 0.25, 0.5, 0.6, 0.7, 0.8,
  0.9, 0.95 and 0.99) of daily mean precipitation

19

• PDFs of daily mean surface air temperature over
  the whole analysis region and the Source Region
  of Yellow River

20

• Quantiles (from 0.05 to 1 in steps of 0.05) of
  daily mean surface air temperature

21
Simulation resultExtreme index
22

• Precipitation extreme index definitions

Variable name Definition
Consecutive dry days (CDD) Maximum number of consecutive days with Precipitation lt 1mm
Number of heavy precipitation days (R10) Annual count of days when Precipitation gt 10mm
Maximum 5-day precipitation amount (Rx5 day) Annual maximum consecutive 5-day precipitation
Very wet days (R95) Annual total precipitation when Pre. gt 95th percentile
23
Extreme precipitation index
24

• Temperature extreme index definitions

Variable name Definition
Summer day (SU) Daily maximum temperature over 25?
Consecutive frost days (CFD) Days with daily minimum temperature below 0?
Growing season length (GSL) The number of days between the first occurrence of at least 6 consecutive days with daily mean temperature above 5? and the first occurrence after 1st July of at least 6 consecutive days with daily mean temperature below 5?
25
Extreme temperature index
26
Summary

1. The RegCM3 model displays wet bias and cold bias
   with a better performance in the Source Region of
   Yellow River. And the wet bias is significantly
   larger in percent during winter
2. The model accurately captures the interannual
   variability and annual cycle of both
   precipitation and temperature averaging over the
   entire region with high correlation coefficients
3. It can also well simulate the probability
   distribution (PDFs) of precipitation but
   underestimate the extreme precipitation in summer
   and overestimate it in winter. The simulated
   temperature PDFs are shifted towards the lower
   temperatures
4. The RegCM3 model generally reproduces the spatial
   patterns of the extreme indices of precipitation
   and temperature but tends to overestimate the
   heavy rainfall and cold days
5. The simulation ability is improved in a great
   degree over Source Region of Yellow River by
   using higher resolution

27

• Thank You!