Fede al Depa men o Home A ai s FDHA
Fede al O ice o Me eo ology and Clima ology Me eoSwiss
1/6
Documen a ion o Me eoSwiss G id-Da a P oduc s
Hou ly Mean Tempe a u e: TabsH
Figu e 1: Hou ly mean empe a u e (°C) o 22. Janua y 2020, 3:00 UTC.
A ypical empe a u e dis ibu ion du ing a basin-scale in e sion o e he Swiss Pla eau.
Va iable
TabsH ep esen s su ace ai empe a u e, 2 m abo e g ound le el, ep esen a i e o he a -
e age o e ci il hou s. Mo e speci ically, he alue a hou HH ep esen s an a e age o six
ins an aneous measu emen s in 10-minu e in e als om (HH – 1):10 o HH:00. Values a e
gi en in deg ees Celsius. An associa ed da a p oduc (TabsHunc) cha ac e izes he s anda d
e o o he es ima ed empe a u es in TabsH, also gi en in deg ees Celsius.
Applica ion
High- esolu ion physico-chemical modeling o he en i onmen (soil wa e , g ound wa e , un-
o , ecosys ems and ag icul u al c ops, snow co e and glacie s). Wea he and clima e e-
la ed economic sec o s ( ou ism, ene gy, cons uc ion). Clima e change downscaling.
O e iew
The g id da ase desc ibes he spa ial dis ibu ion and hou - o-hou e olu ion o su ace ai
empe a u e on a 1-km g id wi hin and a ound Swi ze land. TabsH is cu en ly a ailable o 7
yea s (2018-2024), bu may be ex ended o a longe ime pe iod, depending on in e es and
esou ces. TabsH builds on in-si u measu emen s a app ox. 250 wea he s a ions. I is in-
ended o applica ions ha equi e explici esolu ion o he diu nal cycle.
−10
−5
0
5
Hou ly Mean Tempe a u e (°C) on 22.01.2020 03:00 UTC
2025−10−20
TabsH 1.0
© Me eoSwiss
Hou ly Mean Tempe a u e: TabsH
2/6
Da a base
TabsH elies on measu emen s o nea -su ace ai empe a u e aken a he ope a ional s a-
ion ne wo k SwissMe Ne o Me eoSwiss (Me eoSwiss 2018). I also inco po a es da a om
he nea -bo de sec ions o he s a ion ne wo ks in neighbo ing coun ies. (Da a ob ained
om Geosphe e Aus ia, he Ge man Wea he Se ice DWD, Me eoF ance, and he egional
agencies (ARPAs) o Valle d’Aos a, Piemon e and Lomba dia in I aly). The o al numbe o
s a ions used is 259. The median and minimum numbe o measu emen s a ailable a any
hou a e 258 and 212, espec i ely. Mo e de ail on he s a ion sample o his p oduc , and a
map o he s a ion loca ions, can be ound in F ey & F ei (2025, sec ion 2.2).
Measu emen s in Swi ze land a e aken au oma ically a 10-minu e in e als. Hou ly alues
a e de i ed by a e aging he 10-minu es measu emen s om (hh-1):10 o hh:00 (ci il-hou ly
mean). Measu emen s in Aus ia, Ge many, and in he I alian egions Valle d’Aos a and Lom-
ba dia a e hou ly a e ages as well. In F ance and he egion Piemon e, howe e , hou ly ob-
se a ions we e a ailable only as ins an aneous alues a he ull hou . In hese cases,
pseudo-mean alues ha e been calcula ed, p io o he analysis, by a e aging he wo alues
a he beginning and he end o he hou . The di e ence in empo al ep esen a i i y o e he
hou be ween he di e en s a ion ne wo ks is accoun ed o in he e o model used in he in-
e pola ion.
Tempe a u e measu emen s used in he analysis we e all aken 2 me e s abo e g ound le el,
ollowing he guidelines and s anda ds o WMO (WMO 2008). The measu emen s in Swi ze -
land a e checked igo ously o da a quali y by he ope a ional p ocedu es a Me eoSwiss. A
he o he da a p o ide s, e i ica ion p ocedu es a e in place oo, bu esidual da a quali y is-
sues ha e been encoun e ed in he de elopmen o TabsH. The e o e, addi ional es s we e
conduc ed on he spa ial consis ency o he measu emen s, using a spa ial s a is ical model.
Measu emen s wi h an absolu e de ia ion o >10°C om he model p edic ion we e excluded.
Me hod
The cons uc ion o TabsH builds on a new s a is ical me hod de eloped speci ically o ad-
d ess he challenges o empe a u e in e pola ion a sub-daily ime scales and in complex o-
pog aphy. The me hod is desc ibed in de ail in F ey & F ei (2025). This sec ion p o ides a
sho summa y.
The me hod is based on a spa io- empo al dynamic linea model (DLM). I models he spa ial
and he empo al o ganiza ion in he da a join ly, wi h he ad an age ha in o ma ion in he
da a is also ans e ed o e ime, no jus o e space like in s anda d in e pola ion. This al-
lows o connec pa e ns in he diu nal he mal cycle o pa e ns in space, as is desi able
when pic u ing, o example, he dampened / enhanced he mal a ia ions along lake sho es /
in inne -Alpine alleys. DLM is a concep ual ex ension o k iging wi h ex e nal d i (KED), a
model widely used in spa ial clima ology, bu i in ol es addi ional pa ame e s o cha ac e ize
he empo al dependence o he d i coe icien s.
Fo TabsH, he spa io- empo al DLM is ca e ully con igu ed o ep esen he clima ic cha ac-
e is ics o he egion. A numbe o spa ial co a ia es is included o desc ibe he (occasionally
non-linea ) a ia ion o empe a u e wi h ele a ion. Addi ional spa ial co a ia es a e used o
model local-scale e ec s wi h po en ially la ge ampli udes, such as cold-ai pooling (CAP),
lake- and alley-e ec s, and Foehn. The p edic o se also in ol es a no h-sou h indica o
and wes -eas g adien o accoun o la ge-scale con as s in he empe a u e p o ile and he
local e ec s, such as when an in e sion is p esen o he No h, bu no o he Sou h o he
main Alpine c es , o when oehn is p esen in he Eas bu no in he Wes o he domain. Re-
siduals a e modelled, spa ially, as a ze o-mean Gaussian andom ield wi h an exponen ial
co a iance and a nugge e ec . The empo al a ia ion o he co a ia e coe icien s is mod-
elled as a supe posi ion o a slowly e ol ing end and se e al ha monic diu nal oscilla ions.
The ampli ude and phase o he la e can a y g adually o e ime.
Hou ly Mean Tempe a u e: TabsH
3/6
F ey & F ei (2025) illus a e and e alua e he spa io- empo al DLM, wi h he con igu a ion jus
desc ibed, in wo case s udies wi h challenging condi ions. The me hod is shown o plausibly
ep oduce spa ial a ia ions o he diu nal he mal cycle, including he build-up o a basin-
scale empe a u e in e sion and i s g adual (diu nal) a ia ion o e he cou se o se e al
days. The case s udy also shows ha he model success ully disen angles he spa io- em-
po al pa e ns o he lake-, alley- and CAP-e ec s om he inpu da a.
Tempe a u e p edic ions in a high-moun ain a ea a e subjec o unce ain y ha may be ele-
an o conside in applica ions. The s a is ical model employed o TabsH also in o ms abou
hese unce ain ies. We p o ide his in o ma ion as a sepa a e da ase , deno ed TabsHunc,
on he same g id and a he same ime poin s as TabsH. TabsHunc is a measu e o he
s anda d de ia ion o he dis ibu ion o he p edic ion e o . I should be no ed ha TabsHunc
is abou e o s in he p edic ion o he ue alue and, hence, i is no abou he dis ibu ion o
po en ial measu emen s which would also in ol e measu emen s e o s. An example ield o
TabsHunc is depic ed in Figu e 2. I shows ha he magni ude o he p edic ion e o s ongly
depends on dis ance om measu emen s.
The digi al ele a ion model (DEM) used o cons uc TabsH is om he SRTM (1-km g id,
Fa e al. 2007). Some manual adjus men s ha e been made o a oid alley cu -o s as a e-
sul o he ini e esolu ion. The DEM is a ailable wi h he da ase s.
Figu e 2: TabsHunc o 22. Janua y 2020, 3:00 UTC. The alues quan i y one s anda d de i-
a ion o he p edic ion e o o TabsH, shown in Figu e 1. Values a e gi en in °C.
Ta ge use s
The hou ly ime esolu ion o e ed in TabsH p o ides an explici desc ip ion o he diu nal a i-
a ion o empe a u e, no co e ed in he daily da a p oduc (TabsD). I is he e o e sui able o
all kinds o applica ions ha equi e de ail in he empo al e olu ion. The e is a wide ange o
en i onmen al p ocesses and pe inen modeling applica ions ha depend on sub-daily eso-
lu ion. Examples a e soil-a mosphe e wa e luxes (e.g. e apo a ion, snow/ ain dis inc ion),
he e olu ion o snow co e s (e.g. mel ing), he g ow h o economic plan s (c ops, pas u e),
he mass balance o glacie s, e c. Simila equi emen s a e me in socio-economic sec o s
such as in heal h (hea s ess), sola powe p oduc ion (e iciency o pho o ol aic panels) and
enginee ing (building design). TabsH pe mi s o model he diu nal he mal cycle explici ly
and, hence, allows o a mo e physical a he han pa ame e ized ep esen a ion o p o-
cesses.
The spa ially comple e analysis also p o ides a aluable e e ence o he e alua ion o
model-based empe a u e o ecas s, and high- esolu ion clima e models. Owing o he hou ly
esolu ion, such ac i i ies may also add ess he ep esen a ion o ex eme wea he condi ions
Hou ly Mean Tempe a u e: TabsH
4/6
in nume ical models, such as hea wa es and ela ed soil d ough , he he mal imp in s o
synop ic on s, oehn, e c.
A majo limi a ion o applica ions o TabsH is i s sho empo al ex en . Se en yea s a e a-
he limi ed o applica ions ega ding in e annual clima e a ia ions, clima e moni o ing, o
model e alua ion beyond case s udies. An ex ension o TabsH u he in o he pas would be
equi ed o acili a e such applica ions.
Accu acy and
in e p e a ion
Despi e i s speci ic design o a moun ainous a ea and he compa a i ely dense measu e-
men ne wo k, he es ima es o empe a u e o non-ins umen ed loca ions a e subjec o
e o s, which a e c ucial o conside in applica ions o he da a p oduc s.
G idpoin e sus si e ele a ion: TabsH p o ides empe a u e es ima es o he g id poin s o
a 1-km digi al ele a ion model (DEM), which may di e om he ele a ion o a loca ion o
in e es . Use s needing es ima es o speci ic loca ions o on a di e en DEM will ha e o
co ec o ele a ion di e ences. A e ical in e pola ion using 9 su ounding g id poin s is
mos ly su icien o his. The DEM used o TabsH is a ailable wi h he da a p oduc s.
Un esol ed scales: Se e al small-scale e ec s on he empe a u e dis ibu ion a e no mod-
eled in he analysis. Among hese a e all kinds o land co e e ec s (e.g. lakes and u ban
hea islands) and he in luence o local opog aphy. As a esul , i mus be expec ed ha
spa ial a ia ions a e unde es ima ed ( oo smoo h), pa icula ly a he scale o he g id-poin
spacing, and small-scale pa e ns may display wi h conside able e o in ex en and ampli-
ude, o missed al oge he . An example o such pa e ns a e local-scale wind sys ems, like
he Schöllenen wind in Göschenen, associa ed wi h e y local he mal anomalies o la ge
ampli ude (see e.g. sec ion 4.3 o F ey & F ei 2025).
In e pola ion unce ain y: The in e pola ion accu acy has been calcula ed by lea e-one-ou
c oss alida ion o e a 2.5-yea pe iod. The esul s a e shown in Figu e 3 below. The e o s
ha e p onounced diu nal and seasonal pa e ns, ha a y by geog aphic sub- egion.
In he diu nal cycle (Figu e 3a), he mean absolu e e o (MAE) is gene ally somewha
la ge a nigh han du ing he day. A nigh , he MAE eaches alues o a ound 0.8°C in he
la land (g een solid line in Figu e 3a), 1.0°C in high-moun ain loca ions ( ed and blue solid
lines), and up o 1.2°C in moun ain alleys ha a e s ongly a ec ed by CAP (pu ple solid
line). Du ing he day, he MAE is abou 0.5°C in he la land (g een solid line), 0.8°C a high-
moun ain loca ions ( ed and blue solid lines), and 1.0°C in moun ain alleys (pu ple solid
line). P edic ions gene ally ha e small biases (mean e o s) only. No able excep ions a e
moun ain op loca ions, whe e he diu nal cycle is o e es ima ed ( ed dashed line), and
moun ain alleys, whe e p edic ions a e wa m-biased (pu ple dashed line), due o an un-
de es ima ion o he CAP e ec .
In he annual cycle (Figu e 3b), he mean absolu e e o (MAE) is la ges in win e and
anges om 0.8°C o e he la land (g een solid line) o 1.1°C o e he Alps ( ed solid line).
Pa icula ly la ge e o s a e ound in moun ain alleys (pu ple solid line, MAE up o 1.5°C),
mos ly due o CAP in hese en i onmen s, which is unde es ima ed (wa m bias shown by
he pu ple, dashed line). In all o he seasons, in e pola ion e o s a e smalle and a ain al-
ues a ound 0.6°C o e he la land (g een line), 0.9°C o e he Alps ( ed and blue lines) and
1.0°C in moun ain alleys (pu ple line). Apa om alley loca ions (pu ple, dashed line),
p edic ions ha e small biases only in he annual cycle.
I is impo an o no e ha alues in Figu e 3 desc ibe mean e o s, sugges ing ha TabsH is
qui e accu a e on a e age. In ce ain cases, howe e , he e o s can be mo e subs an ial. In
win e , abou 10% o he absolu e e o s in he c oss- alida ion u n ou o be la ge han
Hou ly Mean Tempe a u e: TabsH
5/6
3.1°C in moun ain alleys and 1.6°C in he la land. In summe , he quan iles a e lowe , abou
2.1°C in alleys and 1.4°C in he la land.
Figu e 3: E o s a is ics o TabsH, es ima ed om a c oss- alida ion. Shown a e he diu nal
(panel a) and annual cycles (panel b) o he mean absolu e e o (solid lines) and he bias
(mean e o , dashed lines). The esul s a e s a i ied acco ding o di e en g oups o s a-
ions, sha ing simila mic oclima ic en i onmen s. G oup “peaks” ( ed) encompasses s a-
ions si ua ed on moun ain ops a high and medium ele a ions, ha a e mos ly in luenced
by he ee oposphe e. G oup “exposed” (blue) con ains s a ions on slopes and idges a
high and medium ele a ions, also s ongly exposed o la ge -scale winds. G oup “ la ”
(g een) encompasses s a ions a low ele a ions, su ounded by la and hilly e ain, mos ly
in he Swiss Pla eau. Finally, g oup “local” (pu ple) con ains s a ions on alley loo s and
lake sho es ha a e s ongly a ec ed by local-scale phenomena such as cold-ai pooling
and lake e ec s.
Rela ed
p oduc s
TabsD / TabsM / TabsY / TminD / TminM / TminY / TmaxD / TmaxM / TmaxY: Simila o
hou ly mean empe a u es bu o he daily / mon hly / yea ly mean and ex emes o 2m-
empe a u e. (See F ei 2014 o me hod de ail.) The e is no s ic consis ency be ween he
da ase s in he sense ha a e aging hou ly analyses does no exac ly ep oduce daily,
mon hly o yea ly analyses. I daily/mon hly/yea ly esolu ion is su icien , i is ecom-
mended o wo k wi h he analyses o coa se ime esolu ion.
CombiP ecip: Simila o hou ly mean empe a u es, bu o p ecipi a ion.
G id s uc u es
TabsH is a ailable in he ollowing g id s uc u es:
ch01h.swiss.l 95 (1-km g id, swiss coo dina es, domain o hyd ological Swi ze land)
an analysis on a 1.1-km g id o COSMO ( o a ed lon-la ) can be p o ided upon eques
Ve sions
Cu en e sion: TabsH 1.0
P e ious e sions: none
Upda e cycle
Cu en ly, he da a se is s a ic. No upda es a e planned a p esen . An ex ension u he
in o he pas , o an ope a ional con inua ion in nea - eal ime may be conside ed in he u-
u e, bu his would equi e esou ces.
Hou ly Mean Tempe a u e: TabsH
6/6
Re e ences
Fa , T. G., Rosen, P. A., Ca o, E., C ippen, R., Du en, R., Hens- ley, S., Kob ick, M., Palle , M., Rod iguez, E., Ro h,
L., Seal, D., Sha e , S., Shimada, J., Umland, J.,We ne , M., Oskin, M., Bu - bank, D., and Alsdo , D.: The
shu le ada opog aphy mission, Re . Geophys., 45, RG2004, doi:10.1029/2005RG000183, 2007.
F ei, C., 2014: In e pola ion o empe a u e in a moun ainous egion using non-linea p o iles and non-Euclidean di-
s ances. In . J. Clima ol., 34, 1585-1605. doi: 10.1002/joc.3786.
F ey, L. and F ei, C. 2025. In e pola ion o Hou ly Su ace Ai Tempe a u e — Applica ion o a Spa io-Tempo al S a i-
s ical Model in Complex Te ain. In e na ional Jou nal o Clima ology 45 (7), 1-20. DOI: 10.1002/joc.8796
Me eoSwiss. 2018. SwissMe Ne : The Me eoSwiss Re e ence Moni o ing Ne wo k. Technical Repo , Me eoSwiss.
WMO, 2008: Guide o me eo ological ins umen s and me hods o obse a ion. Wo ld Me eo ological O ganisa ion
WMO-No. 8, Se en h Edi ion, 681 pp.
No embe 2025
