We

eather Data Playlist

We
eather Data
D
in GBS
G
Goo
od climate data is a critical inp
put for valid
d building performancce analysiss. The Auttodesk Clim
mate
Serrver provide
es easy ac
ccess to a large datab
base of clim
mate and weather
w
datta within Autodesk Grreen
Buillding Studio
o, Revit, an
nd Vasari.
Wea
ather can vary greattly from sitte-to-site, even within a relativ
vely small geographiccal area. Most
M
wea
ather data comes from
m physical weather stations, wh
hich are situated at la
arge airportts and mayy not
reprresent the weather
w
at your buildiing. Getting
g more loca
ally appropriate weath
her data can help improve
the accuracy o
of your ene
ergy analysis results.
Auto
odesks C
Climate Se
erver has 1.6 millio
on virtual weather stations
s
ba
ased on meteorolog
gical
simulations. T
This meanss that you can
c
usually
y find usefful weatherr data less than 14km
m away fro
om a
building site allmost anyw
where in the
e world.
How
w much do
oes weathe
er data ma
atter to ene
ergy simullation resu
ults?
Ene
ergy simula
ation resultts are only as good as
a the unde
erlying model. This in
ncludes bo
oth the building
geo
ometry and the environ
nmental co
onditions lik
ke weather..
You
ur buildingss climate and
a its design will dete
ermine the weather variables that your ene
ergy model will
be the
t most ssensitive to. For exam
mple, if your building h
has a lot of
o glazing and
a is in a sunny clim
mate,
sola
ar radiation
n is likely to
o be a dom
minating facctor. Generally, to ge
et good sim
mulation ressults for HV
VAC
systtems, youlll need accurate data on drybulb
b temperatu
ure and dew
w-point tem
mperature.
Also
o, because
e cooling can
c
be so much morre thermod
dynamicallyy efficient than
t
heatin
ng, change
es to
wea
ather conditions have less of an impact on energy usse results in
n cooling-d
dominated locations
l
(s
such
as Los Angele
es) than in
n heating dominated
d
locations (like Copenhagen). So
S you nee
ed to be extra
e
eful about yyour weather data in cold
c
climate
es.
care
Las
stly, if your w
weather da
ata is generally correcct and conssistently applied, it sho
ould be suffficient for early
e
com
mparative design studies.
Are
e Climate D
Data and Weather
W
Da
ata differen
nt - and wh
hen to use
e each?
Clim
mate data is based on
o historical average
es, usually over a pe
eriod of record (PoR
R) of 30 ye
ears.
Clim
mate data is most usseful for prredictive whole
w
buildiing energyy simulation
ns because
e it represents
typical conditio
ons. That said,
s
typica
al years ma
ay not be a good indicator of the
e condition
ns over a given
yea
ar or of a ch
hanging clim
mate. It doe
es not repre
esent extre
eme conditions (i.e. ab
bnormally h
hot or dry).
Wea
ather data is from a specific
s
da
ay and time
e. Within an
n existing building
b
co
ontext, it ca
an allow yo
ou to
simulate a spe
ecific year and comp
pare the re
esults to acctual energy use. This is useful for creatin
ng a
calibrated ene
ergy model for as-bu
uilt buildin
ngs. This data
d
can also
a
be use
ed for pred
dictive building
ergy simulations, with the caveatt that it is more
m
appropriate early
y, compara
ative analyssis when yo
oure
ene
tryin
ng to unde
erstand how
w energy use
u responds to differrent design
n options. It
I is not me
eant for HV
VAC
sizin
ng, for exa
ample. Forr comparative analysiis, the ben
nefit of the close proximity of th
he data to
o the
building site can
c
outweigh the facct that it is for a sing
gle year, an
nd not bassed on long-term clim
mate
servations.
obs

Where is data
a available
e?
The
ere are virtu
ual weathe
er stations available fo
or most of the world. Exceptions include parts
p
of central
Africca, central Australia, Russia, an
nd extreme
e latitudes near
n
the po
oles. Learn
n more abo
out the sources
of th
his data in the Data Sources
S
article.

Parrent topic: Project We

eather Stattion Tab

We
eather Data
D
Sou
urces an
nd Methodologyy
The
e Autodesk Climate Se
erver is a distribution
d
channel that aggrega
ates and de
elivers clima
ate and
wea
ather data ffrom severa
al sources, using seve
eral method
dologies:

1. Virtual we
eather statio
ons ("Green Building Studio wea
ather stations" - 1.6 million
m
statio
ons)
Autodesk Climate Se
erver includ
des about 1.6
1 million vvirtual wea
ather station
ns based on numerica
al
ogical simulations. It was
w created
d to allow more
m
accura
ate simulattions in placces where
meteorolo
there isn'tt a physical weather station.
s

This data comes fro

om simulations of spe
ecific yearss (2004 and 2006). Unlike
U
the TMY2
T
data
a, no
de to select data to be
e more charracteristic of
o the clima
ate. As a re
esult, this is
s not
effort has been mad
ata and any
y extremess or anomallies that occurred in th
hat year are
e reflected in the data
a.
climate da
To know what year the data ccomes from
m, look at the
t name of
o the virtual weatherr station: if it is
"GBS_04"" it is from 2004,
2
for example.
e
Even so, this data may
m be mo
ore appropriate for bu
uilding sitess than the closest ph
hysical wea
ather
n areas of significant
s
terrain and
d elevation variability and when the buildin
ng or
station, esspecially in
TMY station are nea
ar significan
nt bodies off water.
How the s
simulation
ns work:
The data at virtual weather
w
sta
ations come
e from a nu
umerical weather
w
prediction (NW
WP) model that
he atmosph
here and projects
p
the
e weather variables of interest. An
is based on the physics of th
a
of sev
veral hundrred square kilometerss is broken-up into a grid of 10 and 20 squ
uare
analysis area
kilometer segments - and the ssimulation is run within
n each squ
uare of that grid. The resulting
r
virtual
weather sstation is located at the center off each grid square.
The simu
ulations are
e initialized
d with mete
eorologicall data assiimilated fro
om a varie
ety of sourrces,
including: surface re
eporting sta
ations and buoys, com
mmercial aircraft,
a
win
nd profilers, rawinsondes,
ondes, Rad
dio Acoustic Sounding
g System (RASS), Ve
elocity-Azim
muth Displa
ay (VAD) winds
w
dropwinso
from Dopp
pler radar, GOES, GP
PS, and SS
SM/I satellite
e data.
The weath
her prediction modelss used exhibit high skiill at foreca
asting weath
her in the first 48 hours of
the simula
ation perio
od. As the length of the simulation grows, minute diffferences between
b
re
eality
and the initial mod
del conditions are magnified and comp
pounded upon.
u
Afte
er 48-hours
s of
n, the data is added to
o record that will beco
ome the da
ata file for th
he year, an
nd the mod
del is
simulation
re-initializzed with a frresh set of assimilated observations.
The speccific modelss that were
e used to create
c
the Green Building Stud
dio Weathe
er Stations are:
Rapid Up
pdate Cycle
e (RUC) and
a
Mesoscale Meteo
orological Model
M
verssion 5 (MM
M5). RUC was
used for most of th
he data (2004 dataset), but there were some
s
gaps
s in the da
ata. The MM5
M
ns were the
en used to fill
f those ga
aps (2006 dataset).
d
simulation
Why is th
here a wea
ather statio
on in the ocean with a non-zero
o elevation
n?
Because the climate
e simulation
ns are done on a grid
d, the eleva
ations are averaged
a
w
within
that grid.
g
Near the coast, the center of the analysiss square might
m
be in the water - and some
etimes you see
he coast.
the effectss of an elevvated land mass on th
More in
nformation on
o RUC
More in
nformation on
o MM5
Note: Thiss data is labeled as "2
2004/2006"" within Auttodesk Vassari and Revit.
See this w
whitepaper for more in
nformation, and a valid
dation stud
dy that was done in 20
008.

2. Physical weather stations ("other weather stations" - about 1,800 stations)

Autodesk Climate Server includes about 1,800 stations with TMY2 data. This data comes from
measured values at weather stations, which are often either at airports or maintained by other
groups.
The National Climatic Data Center (NCDC) also maintains databases of TMY data.
More info on TMY2 format
While this data comes from TMY2 data, the output of GBS is not in the TMY2 format. It is different
than the fixed length record format that NREL distributes. GBS drops the non-metrological meta
data (source and uncertainty flags) and converts the units.
Note: This data is labeled as "2010" within Autodesk Vasari and Revit, although it does not
correlate to that specific year because it is climate data. It includes data from 12 different years of
data (selected from the past 30 years).
3. Historical data (3-Tier3)
3-Tier data is simulated data from the last 10-20 years. This reanalysis data is generated by the
Weather Research and Forecasting (WRF) model. It is simulated in a similar way to the virtual
weather stations described above. The WRF model is a newer generation methodology than the
RUC & MM5 data that the other virtual weather stations use.
Just like other numerical weather prediction models, WRF outputs weather variables computed at
distinct grid points. However, the data from 3-Tier is spatially interpolated to provide data at the
specific location of the building being simulated.
This data allows you to look back in time and calibrate an "as-built" energy model of an existing
building. This can help with energy efficiency measures and operations & maintenance.
3-Tier data currently cannot be downloaded, and can't be accessed for general predictive
analysis. It can be accessed if you upload utility data through GBS to analyze and calibrate an
existing building model.
More information on 3Tier
Parent topic: Project Weather Station Tab