scieee Open visual document viewer

Biogeophysical impacts of peatland forestation on regional climate changes in Finland

Gao, Y.,Markkanen, T.,Backman, L.,Henttonen, H.M.,Pietikäinen, J.-P.,Mäkelä, H.M.,Laaksonen, A.

Full text

Biogeosciences, 11, 7251–7267, 2014 www.biogeosciences.ne /11/7251/2014/ doi:10.5194/bg-11-7251-2014 © Au ho (s) 2014. CC A ibu ion 3.0 License. Biogeophysical impac s o pea land o es a ion on egional clima e changes in Finland Y. Gao1,2, T. Ma kkanen1, L. Backman1, H. M. Hen onen3, J.-P. Pie ikäinen1, H. M. Mäkelä1, and A. Laaksonen1,4 1Finnish Me eo ological Ins i u e, P.O. Box 503, 00101 Helsinki, Finland 2Uni e si y o Helsinki, Depa men o Physics, P.O. Box 64, 00014 Helsinki, Finland 3Finnish Fo es Resea ch Ins i u e, P.O. Box 18, 01301 Van aa, Finland 4Uni e si y o Eas e n Finland, Depa men o Applied Physics, P.O. Box 1627, 70211 Kuopio, Finland Co espondence o: Y. Gao ([email p o ec ed]) Recei ed: 11 June 2014 – Published in Biogeosciences Discuss.: 22 July 2014 Re ised: 10 No embe 2014 – Accep ed: 12 No embe 2014 – Published: 17 Decembe 2014 Abs ac . Land co e changes can impac he clima e by in- luencing he su ace ene gy and wa e balance. Na u ally eeless o spa sely eed pea lands we e ex ensi ely d ained o s imula e o es g ow h in Finland o e he second hal o 20 h cen u y. The aim o his s udy is o in es iga e he bio- geophysical e ec s o pea land o es a ion on egional cli- ma e in Finland. Two se s o 18-yea clima e simula ions we e done wi h he egional clima e model REMO by us- ing land co e da a based on p e-d ainage (1920s) and pos - d ainage (2000s) Finnish na ional o es in en o ies. In he mos in ensi e pea land o es a ion a ea, loca ed in he mid- dle wes o Finland, he esul s show a wa ming in Ap il o up o 0.43K in mon hly-a e aged daily mean 2m ai em- pe a u e, whe eas a sligh cooling om May o Oc obe o less han 0.1K in gene al is ound. Consequen ly, snow clea - ance days o e ha a ea a e ad anced up o 5 days in he mean o 15yea s. No clea signal is ound o p ecipi a- ion. Th ough analysing he simula ed empe a u e and en- e gy balance e ms, as well as snow dep h o e i e selec ed sub egions, a posi i e eedback induced by pea land o es a- ion is ound be ween dec eased su ace albedo and inc eased su aceai empe a u ein he snow-mel ing pe iod. Ou mod- elled esul s show good quali a i e ag eemen s wi h he ob- se a ional da a. In gene al, dec eased su ace albedo in he snow-mel ing pe iod and inc eased e apo anspi a ion in he g owing pe iod a e he mos impo an biogeophysical as- pec s induced by pea land o es a ion ha cause changes in clima e. The esul s om his s udy can be u he in eg ally analysed wi h biogeochemical e ec s o pea land o es a ion o p o ide backg ound in o ma ion o adap ing u u e o es managemen o mi iga e clima e wa ming e ec s. Mo eo e , hey p o ide insigh s abou he impac s o p ojec ed o es a- ion o und a a high la i udes due o clima e change. 1 In oduc ion Clima e esponse o an h opogenic land co e change hap- pens mo e locally and occu s on a much sho e ime scale compa ed o global wa ming due o inc eased g eenhouse gases (GHG) (IPCC, 2013). The in luences on he cli- ma e om he biogeophysical e ec s caused by land co e changes can enhance o educe he p ojec ed clima e change (Ba hiany e al., 2010; Bonan, 2008; Feddema e al., 2005; Gálos e al., 2011; Gö el e al., 2008; Ge and Zou, 2013; Pielke e al., 2011, 1998; Pi man, 2003). Especially o he clima e impac s o pas la ge-scale a o es a ion, s udies show ha he mos ob ious e ec s o he inc ease o o es s in bo eal a eas a e wa ming du ing snow-co e pe iods due o dec eased su ace albedo and cooling in summe ime om inc eased e apo anspi a ion (ET) in opical a eas wi h su i- cien soil mois u e (Bala e al., 2007; Be s, 2000; Be s e al., 2007). Vas a eas o na u ally eeless o spa sely eed pea - lands ha e been d ained o g ow o es s o imbe p o- duc ion in no he n Eu opean coun ies (Päi änen and Hånell, 2012). In Finland, i is he dominan land co e change o e he las hal cen u y due o he high ac- ion o p is ine pea land and he need o imbe p oduc- ion. The o al pea land a ea o Finland was es ima ed o Published by Cope nicus Publica ions on behal o he Eu opean Geosciences Union. 7252 Y. Gao e al.: Biogeophysical impac s o pea land o es a ion on egional clima e changes in Finland be 9.7millionha in he 1950s (Il essalo, 1956). In he be- ginning o he 2000s, he a ea o d ained pea land o o es y was es ima ed o be 5.7millionha by Minkkinen e al. (2002) and 5.5millionha by Tomppo e al. (2011). The a ea o d ained pea lands is unlikely o inc ease u - he because no mo e public subsidisa ion is gi en o he i s - ime d ainage o pea lands and he inc eased awa eness o na u al conse a ion (Me sä alouden kehi - ämiskeskus Tapio, 1997). The a ea o es o ed mi es was 15000ha be ween 1990 and 2008 (h p://www.biodi e si y. i/en/indica o s/mi es/mi17-mi e- es o a ion) (Kaakinen and Salminen, 2006). Howe e , land co e change is no only a esul o human land-use ac i i ies bu can also be a con- sequence o clima e change. Global wa ming in he u u e is also conside ed o be a ac o ha a ec s bo eal pea land h ough wa e -le el d awdown due o inc eased ET (Laiho e al., 2003; Laine e al., 1995). A en ion has been paid o he clima e e ec s o pea land o es a ion. A dec ease in he local nigh - ime minimum em- pe a u e du ing he g owing season was obse ed oughly o he i s 15yea s a e d ainage (Solan ie, 1994). The eason o his noc u nal cooling phenomenon is he insula ion o lowe soil laye s om he a mosphe e by d y pea . The e- o e, he hea lux om d ained pea soil can no compensa e o he adia i e cooling a he su ace, which leads o a d op in daily minimum empe a u e (Venäläinen e al., 1999). On alonge ime scale, heg owing o es on o me lyopen pea - lands leads o a dec ease in su ace albedo. The easons o his a e he da ke ee co e in compa ison o he ligh e moss/g ass co e in he snow- ee pe iod and he pa ial snow co e in o es a eas compa ed o he ull snow co e in open a eas in he snow-co e pe iod. This inc eases he daily max- imum empe a u e due o an inc ease in he abso p ion o sho -wa e adia ion (Solan ie, 1994). Consis en esul s on he seasonal cycles o su ace albedo and ne su ace so- la adia ion due o pea land o es a ion we e ound by Lo- hila e al. (2010), based on measu emen da a a wo pai s o d ained and und ained pea land si es loca ed in he sou h and no h o Finland. The esul s showed a no ably dec eased su ace albedo and co esponding inc eased ne su ace sola adia ion in sp ing ime. Fu he mo e, Lohila e al. (2010) in- dica ed he local clima e impac s o pea land o es a ion by in es iga ing long- e m (1961–2008) sp ing su ace empe - a u e ends o e sou he n (<65◦N) and no he n (>65◦N) Finland. The la ges posi i e day ime maximum empe a u e end o 0.64Kdecade−1happened in Ap il in sou he n Fin- land, whe e a o al o 2.7millionha o pea lands we e d ained (Hökkä e al., 2002). The nigh - ime minimum empe a u e end h ough he same pe iod was 0.37Kdecade−1. Lohila e al. (2010) a ibu ed he subs an ially la ge inc ease in he day ime maximum empe a u e han in he nigh - ime mini- mum empe a u e o he change in su ace adia i e p ope - ies a e d ainage. Howe e , hese s udies abou he e ec s o pea land o es a ion on clima e a e based on si e-le el da a o obse a ion-based egional da a, which can no a ibu e he clima e impac s o di e en in luencing ac o s. Speci ically, hey can no dis inguish he local biogeophysical e ec s om he global clima e change due o he inc ease in GHG con- cen a ions. The clima e e ec s o pea land o es a ion ha e no been quan i ied on a egional scale/coun y le el o in- es iga e he biogeophysical e ec s in pa icula . Also, he magni ude and pa e n o land-use change e ec s on clima e depend on egional condi ions such as soil p ope y, opog- aphy, e c. In o ma ion om egional s udies is essen ial o he de elopmen o u u e s a egies o clima e mi iga ion o o es managemen . Thus, i is necessa y o in es iga e he e ec s egionally and sys ema ically. In ecen yea s, egional clima e models ha e become sui - able o simula ing egional clima e in a ine esolu ion o esol e small-scale a mosphe ic ci cula ion (Déqué e al., 2005; Jacob e al., 2001, 2007; McG ego , 1997). Fo his, a egional clima e model wi h a ealis ic land scheme o in- e p e mo e de ailed land su ace in o ma ion needs o be applied. In his s udy, he long- e m clima e e ec s caused by pea - land o es a ion a e assessed om wo se s o 15-yea simu- la ion esul s wi h he egional clima e model REMO, by us- ing he his o ical (1920s) and p esen -day (2000s) land co e condi ions. The in en ion o his s udy is o unde s and how pea land o es a ion in Finland in luences egional clima e condi ions h ough biogeophysical p ocesses. 2 Model desc ip ion and me hodology 2.1 REMO clima e model The egional clima e model REMO is a h ee-dimensional hyd os a ic a mosphe ic ci cula ion model de eloped a he Max Planck Ins i u e o Me eo ology in Ge many (Jacob e al., 2001, 2007; Jacob and Podzun, 1997). I s dynamical co e is based on he “Eu opa-Modell”, he o me nume i- cal wea he p edic ion model o he Ge man Wea he Se - ice (Majewski, 1991). The land su ace scheme (LSS) o REMO mainly ollows ha o he global a mosphe e ci cu- la ion model ECHAM4 (Roeckne e al., 1996) wi h se - e al physical package upda es (de ails a e shown below). The p ognos ic a iables a e p essu e, empe a u e, ho izon- al wind componen s, speci ic humidi y, cloud liquid wa e and ice. REMO is d i en by la ge-scale o cing da a acco d- ing o he elaxa ion scheme (Da ies, 1976). The eigh ou e - mos g id boxes a each la e al bounda y a e he sponge zone. Because land co e is cen al o his s udy, a b ie in oduc- ion o he LSS in REMO is gi en below. In REMO LSS, he o al a ea o each model g id box is composed o ac ions o land ( ege a ion co e and ba e soil), wa e (ocean su - ace and inland lake) and sea ice (Semmle e al., 2004). The biogeophysical cha ac e is ics o majo land co e classes (Olson, 1994a, b) a e desc ibed by he ollowing su ace Biogeosciences, 11, 7251–7267, 2014 www.biogeosciences.ne /11/7251/2014/ Y. Gao e al.: Biogeophysical impac s o pea land o es a ion on egional clima e changes in Finland 7253 pa ame e s: backg ound su ace albedo (albedo o e snow- ee land a eas), oughness leng h, ac ional g een ege a- ion co e , lea a ea index (LAI; one-sided g een lea a ea pe uni g ound a ea), o es a io ( ; ac ional co e age o ees ega dless o hei pho osyn he ic ac i i y), soil wa e - holding capaci y (maximum amoun o wa e ha plan s may ex ac om he soil be o e wil ing begins) and olume ic wil ing poin (pe cen age o mois u e in a soil column be- low which plan s s a o wil ) (Hagemann, 2002; Hagemann e al., 1999). The land su ace pa ame e s a e a e aged lin- ea ly acco ding o ac ional co e age o land co e ypes wi hin a model g id box, excep o he oughness leng h ha isa e agedloga i hmically (Claussene al., 1994; Hagemann e al., 1999). As LAI, ac ional g een ege a ion co e and backg ound su ace albedo s ongly depend on he ege a- ion phenology, hey a e p esc ibed wi h in a-annual cycles by using a mon hly a ying g ow h ac o ha de e mines he seasonal g ow h cha ac e is ics o he ege a ion (Hagemann, 2002; Rechid and Jacob, 2006). The g ow h ac o o la i- udes highe han 40◦no h o sou h is de i ed om a 2m empe a u e clima ology (Lega es and Willmo , 1990); in o he la i udes, he ac ion o pho osyn he ically ac i e a- dia ion is used. The simple bucke scheme (Manabe, 1969) is used o soil hyd ology whe e he pa i ioning o su ace uno and in il a ion ollows he A no scheme (Dümenil and Todini, 1992). The soil empe a u e p o ile om he g ound su ace o a ound 10m deep is desc ibed by i e soil laye s wi h in- c easing hickness. The hea conduc i i y and hea capaci y, equi ed in he hea conduc ion equa ion o calcula ing he soil empe a u e, depend on he soil ypes (Ko la ski, 2007). The dis ibu ion o soil ypes is om he FAO/UNESCO soil map o he wo ld (FAO/UNESCO, 1971–1981; Ko la ski, 2007). The A no scheme used o he soil hyd ology was u - he imp o ed by conside ing he high esolu ion subg id- scale he e ogenei y o he ield capaci ies wi hin a clima e model g id box (Hagemann and Ga es, 2003). The esolu ion o subg id-scale he e ogenei y is se o be 10 imes highe han he model esolu ion when using he de aul REMO land co e map-Global Land Co e Cha ac e is ics Da abase (GLCCD) (Lo eland e al., 2000; US Geological Su ey, 2001). The h ee pa ame e s in he imp o ed A no scheme accoun o he shape o he subg id dis ibu ion o soil wa e capaci ies (Be a), subg id minimum (Wmin)and maximum (Wmax)soil wa e capaci ies. Also, he o iginal annual back- g ound albedo cycle was modi ied by using MODIS sa elli e da a be ween 2001 and 2004 in o de o de i e mo e eal- is ic global dis ibu ions o pu e soil albedo and pu e eg- e a ion albedo, which a e hen used o compu e he annual backg ound albedo cycle wi h mon hly a ying LAI (Rechid, 2008; Rechid e al., 2009). Figu e 1. O og aphy o he model domain and he i e selec ed sub- egions (sub egion1 – blue; sub egion2 – ed; sub egion3 – pu ple; sub egion4 – g een; sub egion5 – o ange). The inne black ame shows he ex en o he elaxa ion zone om he ou e bounda y, i.e. he eigh ou e -mos g id boxes in each di ec ion o he model domain. 2.2 The model domain and land co e da a se s Ou model domain co e s Fennoscandia, a pa o Russia and he no he n pa o cen al Eu ope, and i is cen ed on Finland (Fig. 1). Typical ea u es in luencing he clima e o his domain include he No h A lan ic Ocean and he Bal ic Sea ha su ound he Fennoscandian coun ies, many inland lakes loca ed in Sweden and Finland and he ela i ely high Scandina ian moun ain ange; he es o he a ea has a o- pog aphy lowe han 300m abo e sea le el. The de aul land co e map in REMO is he GLCCD. Howe e , i s desc ip ion o he land co e in Finland is un- ealis ic. Fo ins ance, he e is no pea land in Finland in he GLCCD, whe eas 7.4% (22377km2) o he land is co e ed by na u ally eeless o spa sely eed pea lands acco ding o he 10 h Finnish na ional o es in en o y (FNFI10) (Ko ho- nen e al., 2013). The GLCCD was he e o e subs i u ed by he mo e ealis ic and up- o-da e CORINE land co e map (CLC; 2006) o he same model domain in Gao e al. (2014), excep o he Russian pa whe e he CLC (2006) is no a ailable. Un o una ely, land co e maps desc ibing he land co e condi ions o Finland be o e he mos in ensi e pe iod www.biogeosciences.ne /11/7251/2014/ Biogeosciences, 11, 7251–7267, 2014 7254 Y. Gao e al.: Biogeophysical impac s o pea land o es a ion on egional clima e changes in Finland Figu e 2. Changes o ac ional co e age o he 10 land co e classes in Finland om he 1920s o he 2000s (FNFI10–FNFI1). o pea land d ainage in he 1960s a e qui e limi ed. Ne e he- less, he da a collec ed in he 1s Finnish na ional o es in- en o y (FNFI1) p o ide he possibili y o acing back he land co e condi ion o Finland in he 1920s (Il essalo, 1927; Tomppo e al., 2010). Also, he FNFI10, a he han he CLC (2006),is adop ed odesc ibe he landco e condi ion o Fin- land in he 2000s, wi h he aim o a oid he unce ain ies in compa ing land co e maps wi h di e en land co e classi- ica ion me hods and di e en spa ial esolu ions. The FNFI1 and FNFI10 land co e maps a e pos -p oduc s ha we e spe- cially p epa ed o his s udy om he espec i e FNFI ield measu emen da a. The de ailed desc ip iono he p ocedu es o de i ing he FNFI1 and FNFI10 land co e maps is shown in Appendix A. The wo FNFI land co e maps a e in 3km esolu ion and include 10 land co e classes ollowing CLC nomencla u e. The ac ional co e age o he 10 land co e classes o e he land a ea o Finland in he 1920s and he changes om he 1920s o he 2000s based on he wo FNFI land co e maps a e as ollows ( ac ional co e age in he 1920s; changes om he 1920s o he 2000s): coni e ous o - es (33.0%; 5.2%); mixed o es (13.5%; −5.7%); b oad- lea ed o es (4.7%; −0.8%); a i icial a eas (0.7%; 4.1%); na u al g asslands (3.4%; −3.4%); pea bogs (14.3%; −5.2%); open spaces (1.5%; −0.1%); ansi ional wood- land/sh ub (18.9%; 4.3%); moo s and hea hland (2.1%; 0.7%); and ag icul u al a eas (8.0%; 0.9%). Regional di - e ences o hose land co e classes can be seen in Fig. 2. In he FNFI maps, he land co e class ”pea bogs” is de- ined as na u ally eeless pea land and pine mi es whe e he s ocking le el is low o he mean heigh o ees is below 5m a ma u i y. The e o e, he shi ing om pea bogs o o es s ep esen s a majo land co e change due o pea land o es a- ion. In addi ion o egional inspec ions, i e sub egions we e selec ed o ep esen di e en land co e change condi ions be ween FNFI1 and FNFI10 (Fig. 1), and he changes o ac- ional co e age o he 10 land co e classes in hose i e sub egions a e gi en in Table 1. This was done o speci i- cally assess he local clima e e ec s o di e en in ensi ies o pea land o es a ion. F om sub egion1 o sub egion4 he e is a dec ease in he educ ion o pea bogs. Sub egion1 and sub egion2 a e wo pea land o es a ion a eas loca ed in he middle and sou h o Finland espec i ely. In sub egion1 and sub egion2 he e we e dec eases in he ac ional co e age o pea bogs o mo e han 20%, and he dec eases we e mainly compensa ed by coni e ous o es . The dec ease in he ac- ional co e age o pea bogs was 2% less in sub egion2 han ha in sub egion1, bu he inc ease in he ac ional co e age o coni e ous o es was 5% highe in sub egion2 han ha in sub egion1. The o al inc ease in he ac ional co e age o o es ypes was abou 16% in bo h sub egion1 and sub e- gion2. Sub egion3 is loca ed in he eas o sub egion1. The e was a 12% dec ease in he ac ional co e age o pea bogs, bu ins ead o an inc ease o o es s, he ac ional co e - age o ansi ional woodland/sh ub inc eased by 14.3%. Sub- Biogeosciences, 11, 7251–7267, 2014 www.biogeosciences.ne /11/7251/2014/ Y. Gao e al.: Biogeophysical impac s o pea land o es a ion on egional clima e changes in Finland 7255 Table 1. Changes o ac ional co e age (%) o he 10 land co e classes om he 1920s o he 2000s (FNFI10–FNFI1) in he i e sub egions. Class Legend Sub egion1 Sub egion2 Sub egion3 Sub egion4 Sub egion5 1 coni e ous o es 13.40 18.03 −2.24 −11.74 −10.13 2 mixed o es 1.23 −3.46 −2.30 −1.86 −2.10 3 b oad-lea ed o es 1.24 0.98 1.68 −0.52 −4.11 4 a i icial a eas 4.44 4.95 2.44 5.69 2.52 5 na u al g asslands −4.41 −2.10 −1.71 −2.82 −1.60 6 pea bogs −22.92 −20.82 −12.60 −3.80 8.64 7 open spaces 0.06 −0.12 −0.11 −0.31 −1.14 8 ansi ional woodland/sh ub 3.64 −0.72 14.26 4.84 9.12 9 moo s and hea hland 0.00 0.00 0.00 0.00 −1.37 10 ag icul u al a eas 3.31 3.26 0.57 10.52 0.17 egion4 is an a ea whe e he mos in ensi e an h opogenic ac i i ies ha e occu ed in he i e sub egions. The e was a 14% dec ease in he ac ional co e age o o es ypes and a 3.8% dec ease in ha o pea bogs, wi h a 5.7% inc ease in he ac ional co e age o a i icial a eas and a 10.5% in- c ease in ha o ag icul u e a eas. Sub egion5 is an a ea wi h an 8.64% inc ease in he ac ional co e age o pea bogs and a 16.3% dec ease in he ac ional co e age o o es ypes. He ein one should no ice ha some unce ain ies may a ise om sampling in he FNFI1 and FNFI10 da a. This ap- plies especially o FNFI1, whe e he dis ance be ween in en- o y lines was as high as 26km. The e o e, sub egions ha a e smalle han 100km×100km may no be su icien o ep esen he ac ual land co e changes spa ially. Howe e , he dynamics o he local e ec s o land co e changes on clima e can no be de ec ed when a e aging clima e signals o e la ge a eas wi h di e se land co e changes. The e o e small sub egions, which co e a ange o land co e change in ensi ies, a e chosen o e lec local clima e impac s due o di e en land co e changes. Mo eo e , he FNFI da a only co e he land su ace in Finland wi hou conside ing inland lakes. The e o e, he land–sea mask in he model domain is adop ed om he CLC (2006). In addi ion, he land co e condi ions o he a ea ou side Finland in he model domain a e he same as hose in Gao e al. (2014), i.e. based on he CLC (2006) and he GLCCD, and hus iden ical in bo h simula ions. In o de o make he land su ace pa ame e s mo e sui - able o his s udy, se e al modi ica ions in REMO LSS we e done. De ails o hose modi ica ions a e documen ed in Ap- pendix B. 3 Expe imen design Two simula ions we e conduc ed wi h he FNFI1 and FNFI10 land co e maps, ep esen ing he land co e con- di ions be o e and a e pea land o es a ion ac i i ies in Fin- land espec i ely. The simula ions we e d i en wi h 6-hou ly la e al bounda y condi ions om ECWMF ERA-In e im e- analysis da a (Simmons e al., 2007) om 1 Janua y 1979 o 31 Decembe 1996. The 18-yea o wa d uns we e p e- ceded by 10-yea (1 Augus 1979–1 Janua y 1990) sim- ula ions in o de o s abilise he deep soil empe a u es and soil mois u es. The las 15 yea s (1 Decembe 1981– 30 No embe 1996) ou o he 18-yea o wa d simula ions we e adop ed o u he analysis. The analysed pe iod s a s om 1 Decembe in o de o keep all 3 win e mon hs con- inuous. The simula ed i s 1.5 yea s we e excluded in o de o minimise he in luences o he ini ial bounda y condi ions on simula ed clima e condi ions, which ha e a much quicke adap a ion speed han deep soil empe a u e. The model g id is in an 18km esolu ion ho izon ally and ex ends o e 27 e ical le els (up o 25km). The model ime s ep was se o 90s and he ime s eps o ou pu a iables a e 6-hou ly o 3- D a iables and hou ly o 2-D a iables. Daily da a co e ing 24h a e p ocessed om 18:00UTC on he p e ious day o 17:00UTC on he cu en day. Fo 6-hou ly da a, 18:00UTC on he p e ious day and 00:00UTC, 06:00 and 12:00UTC on he cu en day we e used o daily alues. Fo his s udy domain, he g owing season and he do mancy season co e he pe iod om May o Oc obe and om No embe o Ap il espec i ely. 4 Resul s The land co e change e ec s on egional clima e condi ions in Finland a e analysed based on he di e ences in clima e a iables be ween he pos -d ainage and p e-d ainage sim- ula ions (FNFI10–FNFI1). This “del a change app oach” is adop ed o elimina e he unce ain ies ela ed o model bias (Gálos e al., 2011; Jacob e al., 2008). 4.1 E ec s on clima e o e Finland The di e ences in mon hly-a e aged daily mean 2m ai em- pe a u e (T2m) a e qui e he e ogeneous empo ally and spa- ially. T2m di e ences a e mos p ominen in sp ing ime and summe ime (Fig. 3). The mos no iceable di e ence in T2m, up o 0.43K, akes place in he mos in ensi e pea land www.biogeosciences.ne /11/7251/2014/ Biogeosciences, 11, 7251–7267, 2014 7256 Y. Gao e al.: Biogeophysical impac s o pea land o es a ion on egional clima e changes in Finland Figu e 3. The 15-yea a e aged di e ences (FNFI10–FNFI1) in mon hly-a e aged daily mean 2m ai empe a u e in sp ing and summe mon hs. o es a ion a ea in he middle wes o Finland in Ap il. The wa ming is also e iden in Feb ua y and Ma ch, wi h di - e ences o 0.2K in his a ea. Howe e , T2m u ns o show a sligh cooling, gene ally less han 0.1K, in a ew pa s o his a ea om May o Oc obe . The e a e also wo egions in no he n Finland ha show opposi e changes compa ed o he pea land o es a ion a ea in he middle wes o Fin- land wi h cooling in he sp ing and wa ming in he g ow- ing season. This is because o dec eased o es co e and in- c eased ac ion o pea bogs in hose wo a eas om FNFI1- o FNFI10-based land co e maps. An inc ease o less han 0.2K is seen in T2m in he sou heas o Finland in July and Augus as well as in he e y sou h o Finland h oughou he g owing season, which is mainly due o he change om mixed o es o coni e ous o es and he inc eased a i icial a eas espec i ely. The 15-yea a e aged mon hly p ecipi a- ion shows only small di e ences, less han 10mmmon h−1, in a ied pa e ns in he model domain om Ap il o Augus (no shown). The snow clea ance day is also an impo an indica o o sp ing ime clima e change a high la i udes (Peng e al., 2013). The e o e, he snow clea ance day o each g id box in Finland is de e mined o he 15yea s. The snow clea ance day is de ined he e as he i s day a e which he o al num- be o snow-co e ed days does no exceed he o al numbe Figu e 4. The 15-yea a e aged di e ences (FNFI10–FNFI1) in he snow clea ance days o e model g id boxes in Finland. o snow- ee days, and he selec ion o his day ends be o e midsumme in a yea . The di e ences be ween he 15-yea a e aged snow clea ance days o he wo simula ions (Fig. 4) show almos he same pa e n as he di e ences in T2m in Ap il (Fig. 3). In he pea land o es a ion a ea in he mid- dle wes o Finland, he snow clea ance days a e mos ly ad- anced by 0.5 o 3 days and, in a ew g id boxes, ad anced by up o 5 days in he 15-yea mean. The wo small a eas in he no h o Finland wi h e e se land co e changes in compa - ison o pea land o es a ion show up o 2-day delays in gen- e al. In he e y sou h o Finland, he snow clea ance days a e also gene ally ad anced in acco dance wi h he wa m- ing seen in T2m, bu delayed in se e al sca e ed g id boxes due o inc eased ac ion o a i icial a eas a he expense o o es s. 4.2 E ec s on clima e o e i e sub egions T2m, p ecipi a ion and se e al closely ela ed clima e a i- ables (su ace albedo, ne su ace sola adia ion, snow dep h, ET) o he i e sub egions we e p ocessed in o 11-day un- ning means o educe he in luence o day- o-day a ia ions. The di e ences be ween he simula ions in each o he e- gionally a e aged clima e a iables we e u he a e aged o e he 15yea s (Fig. 5). The da e in o ma ion he ein (day o yea , DOY) ep esen s he middle con ibu ing day o he 11-day a e aging pe iod. T2m o sub egion1 shows a wa ming o 0.1 o 0.2K om Feb ua y un il he end o Ma ch and an e iden peak o inc ease om ea ly Ap il o ea ly May ( om DOY 95 o DOY 125) ha eaches a maximum o 0.5K in la e Ap il. Biogeosciences, 11, 7251–7267, 2014 www.biogeosciences.ne /11/7251/2014/ Y. Gao e al.: Biogeophysical impac s o pea land o es a ion on egional clima e changes in Finland 7257 Discussion Pape | Discussion Pape | Discussion Pape | Discussion Pape | 0 60 120 180 240 300 360 -0.6 -0.4 -0.2 0 0.2 0.4 0.6 0.8 (e) DOY ET (mm) 0 60 120 180 240 300 360 -0.3 -0.2 -0.1 0 0.1 0.2 0.3 0.4 0.5 0.6 (a) DOY T2m (K) 0 60 120 180 240 300 360 -0.02 -0.02 -0.01 -0.01 0 0.01 0.01 0.02 (b) DOY snow dep h (m) Sub egion1 Sub egion2 Sub egion3 Sub egion4 Sub egion5 0 60 120 180 240 300 360 -0.08 -0.06 -0.04 -0.02 0 0.02 0.04 0.06 (c) DOY albedo 0 60 120 180 240 300 360 -6 -4 -2 0 2 4 6 8 (d) DOY ne su ace sola adia ion(W/m2) 0 60 120 180 240 300 360 -0.3 -0.2 -0.1 0 0.1 0.2 0.3 0.4 0.5 0.6 ( ) DOY p ecipi a ion (mm) Figu e 5. The 15- yea a e aged egional mean di e ences (FNFI10 – FNFI1) in 11 day unning mean o daily mean (a) wo-me e ai empe a u e, (b) snow dep h (p esen ed as equi alen wa e ), (c) su ace albedo, (d) ne su ace sola adia ion, (e) ET and ( ) p ecipi a ion o he i e sub egions. 42 Figu e 5. The 15-yea a e aged egional mean di e ences (FNFI10 – FNFI1) in 11-day unning mean o daily mean (a) 2m ai empe - a u e, (b) snow dep h (p esen ed as equi alen wa e ), (c) su ace albedo, (d) ne su ace sola adia ion, (e) ET and ( ) p ecipi a ion o he i e sub egions. T2m o sub egion2 has he same de elopmen as sub egion1 h oughou he whole yea , bu he wa ming is much smalle and he bigges di e ence, only 0.12K, occu s in he begin- ning o Ap il. This is consis en wi h he di e ences in snow dep h. The snow-co e pe iod in sub egion2 is sho e along wi h an ea lie maximum di e ence in snow dep h. Mo e- o e , hose cha ac e is ics o he di e ences in snow dep hs a e in quali a i e ag eemen wi h he di e ences in su ace albedo because snow is he key ac o ha con ols he su - ace albedo in he snow-co e pe iod. F om he beginning o May o he beginning o Oc obe , T2m shows a cooling o less han 0.1K in sub egion1 and sub egion2 because he cooling caused by ET exceeds he wa ming caused by he sligh ly lowe albedo. The a iabili y o he di e ences in ne su ace sola adia ion in he g owing season is induced by he a i- abili y o cloud co e a he han su ace albedo. In No em- be , Decembe and Janua y, he di e ences in T2m a y in bo h di ec ions. A high la i udes, incoming sola adia ion is qui e small and cloud co e ac ion is high in la e au umn and win e . The e o e, he di e ences in su ace albedo a e no able o induce di e ences in ne su ace sola adia ion in his pe iod. Ins ead, he su ace ai empe a u e is sensi- i e o changes in he long-wa e adia ion balance ha may lead o a mosphe ic ai empe a u e in e sion unde a clea sky, mani es ing i sel as ex eme cold su ace ai empe a- u e. Thus, he a iabili y o he di e ences in cloud co e caused by sho - e m a ia ions in he clima e con ibu es o he a ied di e ences in T2m in his pe iod. The di e ences in T2m o sub egion3 show a wa ming o less han 0.1K om DOY 91 o DOY 120 bu also a wa ming in an e en smalle magni ude h oughou he g owing season. The di e ence in su ace albedo in sub egion3 is close o 0, al hough he di e ence in snow dep h is simila o ha o sub egion2 bu wi h a ime lag o a ound 15 days in he mos in ensi e poin . In sub egion4, he snow dep h shows a qui e small inc ease om he beginning o Janua y un il he end o Ma ch, which is consis en wi h he inc ease in su ace albedo and explains he sligh dec ease o up o 0.1K in T2m om he middle o Feb ua y un il he end o Ma ch. Sub e- gion5 displays he opposi e cha ac e is ics compa ed o sub- egion1 and sub egion2 o all he in es iga ed a iables. The absolu e di e ences in snow dep h o sub egion5 a e smalle han hose o sub egion1 bu la ge han hose o sub egion2. Because sub egion5 is loca ed in he no h o Finland, he bigges di e ence in snow dep h occu s la e han ha o sub- egion1. The magni ude o he maximum di e ences in T2m in he snow-co e pe iod o sub egion5 also lies be ween ha o sub egion1 and sub egion2 and happens la e han ha o sub egion1. The di e ences in T2m in he g owing season depend on he su plus o ene gy balance e ms whe e ET mani es s i - sel as la en hea lux. In gene al, he inc ease o ET in sub- egion2 is sligh ly highe han ha in sub egion1. As a con- sequence, he dec ease o T2m in sub egion2 is sligh ly la ge han ha in sub egion1 du ing he g owing season when he albedo di e ence is qui e small. The dec eased ET and he sligh ly dec eased su ace albedo oge he esul in a sligh wa ming du ing he g owing season in he o he sub egions. The ex en s o wa ming in he o he sub egions ollow he magni udes o he dec eased ET because he di e ences in su ace albedo a e almos he same in he g owing season. P ecipi a ion has highe a iabili y han ET h oughou he yea in he i e sub egions. In gene al, he di e ences in p e- cipi a ion a e much la ge in he g owing season han in he do mancy season, when hey a e close o 0mmday−1. In he g owing season, he inc ease in p ecipi a ion o sub egion1 occu s du ing a longe pe iod and has a la ge magni ude han ha o sub egion2. The e a e sligh inc eases in he p ecipi a- ion in sub egion3 and sub egion4, whe eas he p ecipi a ion o sub egion5 shows a dec easing endency in he g owing season, wi h he bigges di e ences less han 0.2mmday−1. Fu he mo e, he maximum and minimum di e ences o g id poin wise and egionally a e aged 11-day unning mean o T2m o e 15yea s o sub egion1 we e in es iga ed as complemen s o he egionally a e aged 15-yea mean di - e ences (Fig. 6). T2m shows a maximum di e ence in g id poin wise o nea ly 2K in he snow-mel ing pe iod o e he 15yea s, which is 1K highe han he maximum di e ence www.biogeosciences.ne /11/7251/2014/ Biogeosciences, 11, 7251–7267, 2014 7258 Y. Gao e al.: Biogeophysical impac s o pea land o es a ion on egional clima e changes in Finland Discussion Pape | Discussion Pape | Discussion Pape | Discussion Pape | 0 60 120 180 240 300 360 -2 -1.5 -1 -0.5 0 0.5 1 1.5 2 2.5 DOY T2m (K) Mean Min_G id Max_ G id Max_Sub egion Min _Sub egion ∇ Figu e 6. Maximum, minimum and mean di e ences o g idpoin -wise and egionally a e aged 11 day unning mean o daily mean wo-me e ai empe a u e o e 15 yea s in sub egion1. 43 Figu e 6. Maximum, minimum and mean di e ences o g id poin - wise and egionally a e aged 11-day unning mean o daily mean 2m ai empe a u e o e 15yea s in sub egion1. in egionally a e aged T2m o e he 15yea s and 4 imes ha o he 15-yea mean o egionally a e aged T2m. The imings o he h ee kinds o maximum di e ences in sp ing de ia e om each o he by 3 o 10 days. The minimum di e ences show only a small de ia ion be ween he g id poin wise and egional mean alues o e he 15yea s. Du ing he snow- mel ing pe iod, he minimum di e ences o egionally a e - aged T2m is abo e 0, bu no ha o he g id poin wise T2m. The sp ing ime di e ences be ween egional mean and g id poin wise ex emes elucida e ha , e en wi hin one sub egion wi h homogenous cha ac e is ics ela ed o pea land o es a- ion, he sp ing wa ming o T2m is empo ally and spa ially he e ogeneous. This implies ha local e ec s a e mo e p o- nounced han he egional and empo al s a is ics can e eal. Fo he es o he yea , he di e ences be ween he maxi- mum (minimum) o he g id poin wise and egionally a e - aged T2m a e small and o a mo e egional na u e. In he pe- iod be ween No embe and Janua y, he la ge a ia ions o maximum (minimum) T2m a e con ibu ed by he in e sion e ec s due o sho - e m a ia ions in he clima e. Addi ionally, o a mo e ho ough unde s anding o he e- la ionships be ween sp ing wa ming and albedo changes in he snow-co e pe iod due o pea land o es a ion, wo co - ela ion ela ionships we e in es iga ed o e he 15yea s o sub egion1 (Fig. 7). One is be ween he maximum empe a- u e di e ence day (DOY) and he maximum su ace albedo di e ence day (DOY). The o he is be ween he in lec ion day o o al albedo ( he day when su ace albedo jus in- ishes a as dec ease om i s win e ime le el; DOY) and he snow clea ance day (DOY). The maximum empe a u e di - e ence days ma ch o maximum albedo di e ence days in 6yea s, and he es o he yea s gene ally show a delayed maximum empe a u e di e ence day compa ed o he max- imum albedo di e ence day, wi h a maximum de ia ion o 14 days. In gene al, he snow clea ance day co ela es well Discussion Pape | Discussion Pape | Discussion Pape | Discussion Pape | 100 110 120 130 140 150 100 110 120 130 140 150 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 (b) In lec ion poin o o al albedo (DOY) Selec ed snow clea ance day (DOY) 70 80 90 100 110 120 130 70 80 90 100 110 120 130 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 19951996 (a) Maximum empe a u e change (DOY) Maximum o al albedo change (DOY) Figu e 7. (a) Co ela ion be ween maximum empe a u e change day (DOY) and maximum o al albedo change day (DOY); (b) co ela ion be ween in lec ion day o o al albedo ( he day when su ace albedo jus inishes a as dec ease om i s win e ime le el; DOY) and he snow clea ance day (DOY). The plo s show egional means o e sub egion1 o all 15 yea s. 44 Figu e 7. (a) Co ela ion be ween maximum empe a u e change day (DOY) and maximum o al albedo change day (DOY); (b) co - ela ion be ween in lec ion day o o al albedo ( he day when su ace albedo jus inishes a as dec ease om i s win e ime le el; DOY) and he snow clea ance day (DOY). The plo s show egional means o e sub egion1 o all 15yea s. wi h he in lec ion poin o su ace albedo. Fo mos yea s, he di e ences a e less han 6 days, bu 3 yea s show di e - ences up o a ound 20 days. In hose yea s, spo adic snow all wi h a small accumula ed snow dep h canno eally in oduce di e ences in o al su ace albedo o e he sub egion bu in- luences he de e mina ion o he snow clea ance day. 4.3 Rela ionships be ween he changes in biogeophysical aspec s and he impac s on clima e To assess he gene ali y o he causal ela ionships be ween land co e changes and clima e a iables, he spa ial co - ela ions be ween changes in he wo su ace ene gy bal- ance ele an a iables, su ace albedo and ET, and T2m a e in es iga ed. Consequen ly, he spa ial co ela ions be- ween changes in su ace albedo and ET and changes in he su ace pa ame e alues a e also explo ed. The co e- la ions wi h ac ional g een ege a ion co e is no shown Biogeosciences, 11, 7251–7267, 2014 www.biogeosciences.ne /11/7251/2014/ Y. Gao e al.: Biogeophysical impac s o pea land o es a ion on egional clima e changes in Finland 7259 in Fig. 8 because LAI and g een ege a ion a io a e bo h modula ed wi h he mon hly a ying g ow h ac o by he same scheme, and hey a e highly co ela ed (Pea son co - ela ion coe icien , 2=0.984 o Ma ch, 2=0.674 o June). Mon hly means o 15-yea a e aged changes in Ma ch and June a e selec ed o ep esen sp ing ime and summe - ime espec i ely. The changes in T2m a e in acco dance wi h he changes in su ace albedo in Ma ch (Fig. 8a), which is al- mos linea ly co ela ed wi h he changes in LAI (Fig. 8c) and o es a io (Fig. 8e). The changes in T2m a e linea ly co ela ed wi h he changes in ET o e mos o he a ea in June (Fig. 8b). In gene al, he changes in ET a e also co - ela ed wi h he changes in LAI (Fig. 8d), oughness leng h (Fig. 8 ) and o es a io (yea ly cons an , no shown), despi e hein luences om d ough s ha may happen inla e summe . O e all, he changes in su ace albedo and ET a e closely de- penden on he changes in land su ace pa ame e s, which a e induced by he changes in ac ional co e ages o land co e ypes in he i e sub egions (Table 1). The changes in T2m a e mainly modula ed by he changes in su ace albedo and ET in sp ing and summe espec i ely. Some g id boxes loca ed in he sou heas o Finland, whe e mixed o es was subs i u ed by mainly coni e ous o es , show de ia ions in he co ela- ions wi h LAI (ma ked by yellow ci cles in Fig. 8b, c, d). In his a ea, LAI inc eased wi h almos no change in o es a io, which led o a ela i ely smalle dec ease in su ace albedo compa ed o o he a eas wi h he same magni ude o changes in LAI in Ma ch; he ET-induced cooling is ou weighed by he albedo-induced wa ming, which causes a sligh wa ming in June. In he ollowing summe mon hs, July and Augus , he ET-induced cooling ypically ge s smalle because o su - ace wa e limi a ion and consequen wa ming. 5 Discussion 5.1 Biogeophysical impac s o pea land o es a ion on egional clima e Su ace albedo shows a no able dec ease in pea land o es a- ion a eas du ing he snow-co e pe iod and a sligh dec ease in he g owing season, whe eas LAI, oughness leng h, ac ional g een ege a ion co e and o es a io inc ease h oughou he yea a e pea land o es a ion. Those changes lead o an inc ease in sp ing ime T2m, which occu s locally in acco dance wi h he dec ease in su ace albedo. In he g ow- ing season, an inc ease in ET ela ed o he inc eased LAI and ac ional g een ege a ion co e leads o mo e ene gy consumed by la en hea lux han gained by sligh ly lowe albedo. Addi ionally, highe oughness leng h can play a ole by inc easing u bulen mixing and consequen ly he mag- ni udes o u bulen luxes. Thus, he sca e ed di e ences in p ecipi a ion in summe a e con ibu ed o mo e con ec- i e s uc u es, while o he es o he yea he p ecipi a ion is basically con olled by la ge-scale me eo ology. F om he Figu e 8. Spa ial co ela ions be ween (a) changes in mon hly- a e aged daily mean 2m ai empe a u e (T2m) and changes in albedo o Ma ch, (b) changes in T2m and changes in ET o June and ela ionships be ween changes in land su ace pa ame e s in REMO LSS ollowing land co e changes and changes in albedo (c, e) (changes in ET, d, ) in he co esponding mon h. The changes in he g id boxes in selec ed sub egions a e shown wi h colou ed do s (sub egion1 – blue; sub egion2 – ed; sub egion3 – pu ple; sub e- gion4 – g een; sub egion5 – o ange). The g id boxes in yellow ci - cles show he changes in he sou heas a ea o Finland. analysis o he esul s in he i e sub egions, he di e ences in he clima e a iables show ha hei magni udes depend on he ex en o land co e changes, while he imings o he ex emesmos ly depend ongeog aphical loca ions (la i udes) ha de ine he adia ion balance h ough he seasonal cycle. Resul s also illus a e a posi i e eedback induced by pea - land o es a ion be ween lowe su ace albedo and wa me T2m in he snow-mel ing pe iod. The wa ming caused by lowe su ace albedo in he snow-co e pe iod due o mo e o es leads o a quicke and ea lie snow mel ing; mean- while, he su ace albedo is educed and consequen ly he su ace ai empe a u e is inc eased. Addi ionally, he maxi- www.biogeosciences.ne /11/7251/2014/ Biogeosciences, 11, 7251–7267, 2014 7266 Y. Gao e al.: Biogeophysical impac s o pea land o es a ion on egional clima e changes in Finland o es s o Finland. Resul s o he gene al su ey o he o es s o he coun y ca ied ou du ing he yea s 1921–1924), Commu- nica iones ex Ins i u o Quaes ionum Fo es alium Finlandiae 11, Val ioneu os on ki japaino, 1927. Il essalo, Y.: Suomen me sä uosina 1921–24 uosiin 1951–53: kolmeen al akunnan me sien in en oin iin pe us u a u kimus ( he o es s o Finland om 1921–24 o 1951–53. A su ey based on h ee na ional o es in en o ies), Communica iones Ins i u i Fo es alis Fenniae, Finnish Fo es Resea ch Ins i u e, Helsinki, Finland, 47, 277 pp., 1956. IPCC: Clima e Change 2013: The Physical Science Basis. Wo king G oup 1 Con ibu ion o he Fi h Assessmen Repo o he In- e go e nmen al Panel on Clima e Change, Camb idge, UK and New Yo k, NY, USA, 1535 pp., 2013. Jacob, D. and Podzun, R.: Sensi i i y s udies wi h he egional cli- ma e model REMO, Me eo ol. A mos. Phys., 63, 119–129, 1997. Jacob, D., Van den Hu k, B. J. J. M., And ae, U., Elge ed, G., Fo elius, C., G aham, L. P., Jackson, S. D., Ka s ens, U., Köp- ken, Ch ., Lindau, R., Podzun, R., Rockel, B., Rubel, F., Sass, B. H., Smi h, R. N. B., and Yang, X.: A comp ehensi e model in e -compa ison s udy in es iga ing he wa e budge du ing he BALTEX-PIDCAP pe iod, Me eo ol. A mos. Phys., 77, 19–43, 2001. Jacob, D., Bä ing, L., Ch is ensen, O. B., Ch is ensen, J. H., De Cas o, M., Déqué, M., Gio gi, F., Hagemann, S., Hi schi, M., Jones, R., Kjells öm, E., Lende ink, G., Rockel, B., Sánchez, E., Schä , C., Sene i a ne, S. I., Somo , S., Van Ulden, A., and Van den Hu k, B.: An in e -compa ison o egional clima e models o Eu ope: model pe o mance in p esen -day clima e, Clima ic Change, 81, 31–52, doi:10.1007/s10584-006-9213-4, 2007. Jacob, D., Ko o a, L., Lo enz, P., Moseley, C., and P ei e , S.: Re- gional clima e modeling ac i i ies in ela ion o he CLAVIER p ojec , Idöjá ás, 112, 141–153, 2008. Kaakinen, E. and Salminen, P.: Mi e conse a ion and i s sho his- o y in Finland, in: Finland – land o mi es, The Finnish En i- onmen , 23, edi ed by: Lindholm, T. and Heikkilä, R., Finnish En i onmen Ins i u e, Helsinki, Finland, 229–238, 2006. Køl zow, M.: The e ec o a new snow and sea ice albedo scheme on egional clima e model simula ions, J. Geophys. Res.-A mos., 112, D07110, doi:10.1029/2006JD007693, 2007. Ko honen, K. T., Ihalainen, A., Vii i, H., Heikkinen, J., Hen onen, H. M., Ho anen, J. P., Mäkelä, H., Ne alainen, S., and Pi kä- nen, J.: Suomen me sä 2004–2008 ja niiden kehi ys 1921–2008 ( he o es s o Finland in 2004–2008 and hei de elopmen om 1921 o 2008), Me sä ie een aikakauski ja, 3, Finnish Fo es Re- sea ch Ins i u e, Helsinki, Finland, 269–608, 2013. Ko la ski, S.: A subg id glacie pa ame e isa ion o use in egional clima e modelling, PhD hesis, Uni e si y o Hambu g, Max Planck Ins i u e o Me e ology, Hambu g, Ge many, 2007. Kuusinen, N., Tomppo, E., and Be ninge , F.: Linea unmixing o MODIS albedo composi es o in e subpixel land co e ype albedos, In . J. Appl. Ea h Obs., 23, 324–333, 2013. Laiho, R., Vasande , H., Pen ilä, T., and Laine, J.: Dynamics o plan -media ed o ganic ma e and nu ien cycling ollowing wa e -le el d awdown in bo eal pea lands, Global Biogeochem. Cy., 17, 1053, doi:10.1029/2002GB002015, 2003. Laine, J., Vasande , H., and Laiho, R.: Long- e m e ec s o wa e le el d awdown on he ege a ion o d ained pine mi es in sou h- e n Finland, J. Appl. Ecol., 32, 785–802, 1995. Lega es, D. R. and Willmo , C. J.: Mean seasonal and spa ial a i- abili y in global su ace ai empe a u e, Theo . Appl. Clima ol., 41, 11–21, 1990. Lohila, A., Minkkinen, K., Laine, J., Sa olainen, I., Tuo inen, J.-P., Ko honen, L., Lau ila, T., Tie ä äinen, H., and Laaksonen, A.: Fo es a ion o bo eal pea lands: impac s o changing albedo and g eenhouse gas luxes on adia i e o cing, J. Geophys. Res.- Biogeo., 115, G04011, doi:10.1029/2010JG001327, 2010. Lo eland, T. R., Reed, B. C., B own, J. F., Ohlen, D. O., Zhu, Z., Yang, L., and Me chan , J. W.: De elopmen o a global land co e cha ac e is ics da abase and IGBP DISCo e om 1km AVHRR da a, In . J. Remo e Sens., 21, 1303–1330, 2000. Majewski, D.: The Eu opa-Modell o he Deu sche We e diens , in: ECMWF Semina on nume ical me hods in a mosphe ic mod- els, 2 (Vol.), Reading, UK, 147–191, 1991. Manabe, S.: Clima e and he ocean ci cula ion 1: I. The a mosphe ic ci cula ion and he hyd ology o ea h’s su ace, Mon. Wea he Re ., 97, 739–774, 1969. McG ego , J.: Regional clima e modelling, Me eo ol. A mos. Phys., 63, 105–117, 1997. Me sä alouden kehi ämiskeskus Tapio: Me sä alouden säädökse (Fo es y egula ions), Tapio, Finland, 111 pp., 1997. Mel on, J. R., Wania, R., Hodson, E. L., Poul e , B., Ringe al, B., Spahni, R., Bohn, T., A is, C. A., Bee ling, D. J., Chen, G., Elisee , A. V., Deniso , S. N., Hopc o , P. O., Le enmaie , D. P., Riley, W. J., Singa aye , J. S., Subin, Z. M., Tian, H., Zü che , S., B o kin, V., an Bodegom, P. M., Kleinen, T., Yu, Z. C., and Kaplan, J. O.: P esen s a e o global we land ex en and we land me hane modelling: conclusions om a model in e - compa ison p ojec (WETCHIMP), Biogeosciences, 10, 753– 788, doi:10.5194/bg-10-753-2013, 2013. Minkkinen, K. and Laine, J.: Vege a ion he e ogenei y and di ches c ea e spa ial a iabili y in me hane luxes om pea lands d ained o o es y, Plan Soil, 285, 289–304, doi:10.1007/s11104-006-9016-4, 2006. Minkkinen, K., Ko honen, R., Sa olainen, I., and Laine, J.: Ca bon balance and adia i e o cing o Finnish pea lands 1900–2100 – he impac o o es y d ainage, Glob. Change Biol., 8, 785–799, 2002. Olson, J. S.: Global ecosys em amewo k-de ini ions, USGS EROS Da a Cen e In e nal Repo , Sioux Falls, SD, 37 pp., 1994a. Olson, J. S.: Global ecosys em amewo k- ansla ion s a egy, USGS EROS Da a Cen e In e nal Repo , Sioux Falls, SD, 39 pp., 1994b. Päi änen, J. and Hånell, B.: Pea land Ecology and Fo es y: A Sound App oach, Uni e si y o Helsinki Depa men o Fo es Sciences Publica ion 3, Depa men o Fo es Ecology, Uni e - si y o Helsinki, Helsinki, Finland, 2012. Pebesma, E. J.: Mul i a iable geos a is ics in S: he gs a package, Compu . Geosci., 30, 683–691, 2004. Peng, S., Piao, S., Ciais, P., F iedlings ein, P., Zhou, L., and Wang, T.: Change in snow phenology and i s po en ial eed- back o empe a u e in he No he n Hemisphe e o e he las h ee decades, En i on. Res. Le ., 8, 014008, doi:10.1088/1748- 9326/8/1/014008, 2013. Pielke, R. A., A issa , R., Raupach, M., Dolman, A. J., Zeng, X., and Denning, A. S.: In e ac ions be ween he a mosphe e and e es ial ecosys ems: in luence on wea he and clima e, Glob. Change Biol., 4, 461–475, 1998. Biogeosciences, 11, 7251–7267, 2014 www.biogeosciences.ne /11/7251/2014/ Y. Gao e al.: Biogeophysical impac s o pea land o es a ion on egional clima e changes in Finland 7267 Pielke, R. A., Pi man, A., Niyogi, D., Mahmood, R., McAlpine, C., Hossain, F., Goldewijk, K. K., Nai , U., Be s, R., Fall, S., Re- ichs ein, M., Kaba , P., and de Noble , N.: Land use/land co e changes and clima e: modeling analysis and obse a ional e - idence, Wiley In e disciplina y Re iews: Clima e Change, 2, 828–850, 2011. Pi man, A. J.: The e olu ion o , and e olu ion in, land su ace schemes designed o clima e models, In . J. Clima ol., 23, 479– 510, 2003. Pi man, A. J., de Noble -Ducoud é, N., C uz, F. T., Da in, E. L., Bo- nan, G. B., B o kin, V., Claussen, M., Deli e, C., Ganze eld, L., Gayle , V., an den Hu k, B. J. J. M., Law ence, P. J., an de Molen, M. K., Mülle , C., Reick, C. H., Sene i a ne, S. I., S enge s, B. J., and Voldoi e, A.: Unce ain ies in clima e e- sponses o pas land co e change: Fi s esul s om he LU- CID in e compa ison s udy, Geophys. Res. Le ., 36, L14814, doi:10.1029/2009GL039076, 2009. P euschmann, S.: Regional su ace albedo cha ac e is ics – analysis o albedo da a and applica ion o land-co e changes o a e- gional clima e model, PhD hesis, Uni e si y o Hambu g, Max Planck Ins i u e o Me e ology, Hambu g, 2012. R De elopmen Co e Team, R: A Language and En i onmen o S a is ical Compu ing, The R Founda ion o S a is ical Compu - ing, Vienna, Aus ia, 2011. Räisänen, P., Luoma an a, A., Jä inen, H., Takala, M., Jylhä, K., Bulygina, O. N., Riihelä, A., Laaksonen, A., Koskinen, J., and Pulliainen, J.: E alua ion o No h Eu asian snow-o da es in he ECHAM5.4 a mosphe ic GCM, Geosci. Model De . Discuss., 7, 3671–3715, doi:10.5194/gmdd-7-3671-2014, 2014. Rechid, D.: On biogeophysical in e ac ions be ween ege a ion phe- nology and clima e simula ed o e Eu ope, PhD hesis, Uni e - si y o Hambu g, Max Planck Ins i u e o Me e ology, Hambu g, 2008. Rechid, D. and Jacob, D.: In luence o mon hly a ying ege a ion on he simula ed clima e in Eu ope, Me eo ol. Z., 15, 99–116, 2006. Rechid, D., Radda z, T. J., and Jacob, D.: Pa ame e iza ion o snow- ee land su ace albedo as a unc ion o ege a ion phenology based on MODIS da a and applied in clima e modelling, Theo . Appl. Clima ol., 95, 245–255, 2009. Roeckne , E., A pe, K., Beng sson, L., Ch is oph, M., Claussen, M., Dümenil, L., Esch, M., Gioge a, M., Schlese, U., and Schul z-Weida, U.: The A mosphe ic Gene al Ci cula ion Model ECHAM4: Model Desc ip ion and Simula ion o he P esen - Day Clima e, MPI Repo No. 218, Max Planck Ins i u e o Me- e ology, Hambu g, Ge many, 90 pp., 1996. Roesch, A., Wild, M., Gilgen, H., and Ohmu a, A.: A new snow co e ac ion pa ame iza ion o he ECHAM4 GCM, Clim. Dynam., 17, 933–946, 2001. Semmle , T., Jacob, D., Schlünzen, K. H., and Podzun, R.: In luence o sea ice ea men in a egional clima e model on bounda y laye alues in he F am S ai Region, Mon. Wea he Re ., 132, 985–999, 2004. Simmons, A., Uppala, S., Dee, D., and Kobayashi, S.: ERA-In e im: new ECMWF eanalysis p oduc s om 1989 onwa ds, ECMWF newsle e , 110, 25–35, 2007. Solan ie, R.: Albedo in Finland on he Basis o Obse a ions on Ai c a , Me eo ological publica ions, 12, Finnish Me e ological Ins i u e, Helsinki, Finland, 106 pp., 1988. Solan ie, R.: Suu en suo-oji us en aiku us ilman lämpö ilaan e i- yises i Alajä en Möksyn ha ain ojen pe us eella ( he impac o la ge scale we land d ainage on ai empe a u e based on obse - a ions in Möksy in Alajä i), Me eo ological publica ions, 29, Finnish Me e ological Ins i u e, Helsinki, 40 pp., 1994. Tomppo, E., Gschwan ne , M., Law ence, M., and McRobe s, R. E.: Na ional Fo es In en o ies, Pa hways o Common Repo - ing, Sp inge , The Ne he lands, 2010. Tomppo, E., Heikkinen, J., Hen onen, H. M., Ihalainen, A., Ka ila, M., Mäkelä, H., Tuomainen, H., and Vainikainen, N.: Designing and Conduc ing a Fo es In en o y-case: 9 h Na ional Fo es In- en o y o Finland, Sp inge , The Ne he lands, 2011. Tu e sky, M. R., Ko owska, A., Bubie , J., Dise, N. B., C ill, P., Ho nib ook, E. R. C., Minkkinen, K., Moo e, T. R., Mye s- Smi h, I. H., Nykänen, H., Ole eld , D., Rinne, J., Saa nio, S., Shu pali, N., Tui ila, E.-S., Wadding on, J. M., Whi e, J. R., Wickland, K. P., and Wilmking, M.: A syn hesis o me hane emissions om 71 no he n, empe a e, and sub opical we lands, Global Change Biol., 20, 2183–2197, doi:10.1111/gcb.12580, 2014. US Geological Su ey: Global land co e cha ac e is ics da a base e sion 2.0, a ailable a : h p://edc2.usgs.go /glcc/globdoc2_0. php, 2001. Venäläinen, A., Ron u, L., and Solan ie, R.: On he in luence o pea land d aining on local clima e, Bo eal En i on. Res., 4, 89– 100, 1999. Wiscombe, W. J. and Wa en, S. G.: A model o he spec al albedo o snow. I: Pu e snow, J. A mos. Sci., 37, 2712–2733, 1980. W amneby, A., Smi h, B., and Samuelsson, P.: Ho spo s o ege a ion clima e eedbacks unde u u e g eenhouse o c- ing in Eu ope, J. Geophys. Res.-A mos., 115, D21119, doi:10.1029/2010JD014307, 2010. Zhang, W., Jansson, C., Mille , P. A., Smi h, B., and Samuelsson, P.: Biogeophysical eedbacks enhance he A c ic e es ial ca - bon sink in egional Ea h sys em dynamics, Biogeosciences, 11, 5503-5519, doi:10.5194/bg-11-5503-2014, 2014. www.biogeosciences.ne /11/7251/2014/ Biogeosciences, 11, 7251–7267, 2014