Transcript Arctic Land Surface Hydrology: Moving Towards a Synthesis

Arctic Land Surface
Hydrology:
Moving Towards a Synthesis
Global Datasets
Available Datasets





ERA-40 Reanalysis
NCEP-NCAR Reanalysis
Remote sensing data
Global Runoff Data Center (GRDC, UNH)
Global River Discharge Database
(RivDis, UNH)
 Adam et al. (2006) Precipitation Dataset
 Sheffield et al. (2006) 50-yr
Meteorological Forcings
Global Forcing Dataset
Reanalysis
Observations
Bias-Corrected
High temporal/low
spatial resolution
Generally low temporal/high
spatial resolution
High temporal/high
spatial resolution
CRU
1901-2000, Monthly, 0.5deg
P, T, Tmin, Tmax, Cld
GPCP
1997-, Daily, 1.0deg
P
NCEP/NCAR Reanalysis
UW
PGF50
1948-, 3hr, 6hr, daily, T62
P, T, Lw, Sw, q, p, w
1979-2000, Daily, 2.0deg
P
1948-2000, 3hr, daily, 1.0deg
P, T, Lw, Sw, q, p, w
TRMM
2002-, 3hr, 0.25deg
P
SRB
1985-2000, 3hr, 1.0deg
Lw, Sw
Global Forcing Dataset: Correction of Daily Precipitation Statistics
• High latitude anomaly in
reanalysis rain days
• Corrected to match observed
wetwet, drydry statistics
• By resampling wet and dry
days from reanalysis record
• Other variables resampled
for the same days for
consistency
• Monthly P totals scaled to
match observations
Global Forcing Dataset: Interpolation and Elevation Corrections
• disaggregated from 2.0 to
1.0 degree using bilinear
interpolation but with
adjustments for differences in
elevation between the two
grids
• air temperature adjusted
using the environmental lapse
rate (6.5 oC/km)
• adjust q, p, Lw via water
vapor state equations and
Stefan-Boltzmann law
Difference in elevation between reanalysis and 1.0deg grid
• assumes that the relative
humidity is constant to avoid
the possibility of supersaturation
Global Forcing Dataset: Disaggregation of Precipitation
Disaggregation in Space
p( I | A) p( A)
p( I )
• A = sub-grid area of precipitation
• I = daily precipitation amount
• Bayes theorem used to derive the sub-grid
areal coverage of precipitation for a given grid
precipitation and season
• weighted by neighboring cells
2.0 degree
1.0 degree
• disaggregated from daily to 3-hr by
resampling from TRMM p(3hr|daily)
0.4
Probability of precipitation
p( A | I ) 
Disaggregation in Time
0.35
0.3
0.25
0.2
0.15
0.1
0.05
0
1
2
3
4
5
3-hr period
6
7
8
Global Forcing Dataset: Correction of Radiation
Correction of Sw Trends
• Spurious trend in reanalysis Sw
• Form regression between reanalysis Sw
and Cld
• New Sw time series generated from Cru cld
Monthly Bias Correction of Lw and Sw
• Sw scaled to match SRB
• Lw scaled to match SRB using probability
matching.
Global Forcing Dataset: P, T Monthly Bias Corrections
Precipitation
• P scaled to match observed monthly totals
• Corrected for gauge undercatch
• Orographic corrections can be added
Temperature
• T scaled to match observed monthly totals
• Tmin, Tmax scaled to match observed DTR
Global Retrospective Hydrology Simulations
DJF
MAM
JJA
SON
Mean seasonal relative saturation
Global Retrospective Hydrology Simulations
DJF
MAM
JJA
SON
Mean seasonal evapotranspiration
Global VIC Simulations
Before Calibration
After Calibration
Global Runoff Data Center
 Gridded data at 30-min spatial resolution
 Monthly climatological mean runoff based on
model output and adjusted to match
observations
Global River Discharge
Database
 Data generally from
1969-1984
(http://www.rivdis.sr.unh.edu/)
Adam et al. (2006) Precipitation
 Gridded monthly
precipitation, 1979-99
 Half degree resolution
 Applicable for regions
with high-quality, longterm streamflow data
with few anthropogenic
effects. Basins must
cover area with
orographic effects
(Adam et al., 2006)