Spatio-temporal analysis of remotely sensed and hydrological model soil moisture in the small Jicinka River catchment in Czech Republic

J. Hyd ol. Hyd omech., 69, 2021, 1, 1–12 ©2021. This is an open access a icle dis ibu ed DOI: 10.2478/johh-2020-0038 unde he C ea i e Commons A ibu ion ISSN 1338-4333 NonComme cial-NoDe i a i es 4.0 License 1 Spa io- empo al analysis o emo ely sensed and hyd ological model soil mois u e in he small Jičinka Ri e ca chmen in Czech Republic Vesna Đukić1*, Ranka E ić1, Mi osla Dumb o sky2, Ve onika Sobo ko a2 1 Uni e si y o Belg ade, Facul y o Fo es y, Depa men o Ecological Enginee ing o Soil and Wa e Resou ces P o ec ion, Kneza Višesla a, 1, 11000 Belg ade, Se bia. 2 B no Uni e si y o Technology, Facul y o Ci il Enginee ing, Ins i u e o Landscape Wa e Managemen , An onínská 548/1, 601 90 B no, Czech Republic. * Co esponding au ho . E-mail: [email p o ec ed] Abs ac : The knowledge o spa io- empo al dynamics o soil mois u e wi hin he ca chmen is e y impo an o ain- all– uno modelling in lood o ecas ing. In his s udy he compa ison be ween emo ely sensed soil mois u e and soil mois u e es ima ed om he SHETRAN hyd ological model was pe o med o small and lashy Jičinka Ri e ca chmen (75.9 km2) in he Czech Republic. Due o a ela i ely coa se spa ial esolu ion o sa elli e da a, he sa elli e soil mois u e da a we e downscaled, by applying he me hod de eloped by Qu e al. (2015). The sub-g id a iabili y o soil mois u e was es ima ed on he basis o he mean soil mois u e o he g id cell and he known hyd aulic soil p ope ies. The SHETRAN model was calib a ed and e i ied o he obse ed s eam low hyd og aphs a he ca chmen ou le . The good co ela ion be ween he wo di e en soil mois u e in o ma ion was ob ained acco ding o he majo i y o applied c i e ia. The esul s o he e alua ion c i e ia indica e ha he downscaled emo ely sensed soil mois u e da a can be used as addi ional c i e ia o he calib a ion and alida ion o hyd ological models o small ca chmen s and can con ibu e o a be e es ima ion o pa ame e s, o educe unce ain ies o hyd ological models and imp o e uno simula ions. Keywo ds: SHETRAN hyd ological model; Downscaled emo ely sensed soil mois u e; Runo and soil mois u e alida ion; Spa io- empo al a iabili y o soil mois u e; Flash loods; Small ca chmen . 1 INTRODUCTION The la ge loods ha ha e occu ed in ecen yea s in many egions o he wo ld made local, na ional and in e na ional au ho i ies inc easingly awa e o lood and inunda ion haza d and he necessi y o a be e unde s anding o loods and lood p o ec ion managemen imp o emen (IPCC, 2012). Se e al lash loods occu ed in he e i o y o he Czech Republic du ing he las decade o June and a he beginning o July 2009, and in May and June 2010 (Danhelka e al., 2014). The hilly and lashy Jičinka Ri e basin (75.9 km2) in he Mo a ian - Silesian egion in he Czech Republic was pa icula ly a ec ed by se ious looding due o s eep slopes o he e ain, and a high pe cen age o soil ypes wi h low-in ensi y in il a ion (Pa lik and Dumb o ský, 2014). Rain all– uno models can be e y use ul o lash lood o ecas ing. The uno gene a ing mechanisms a e highly dependan on soil wa e con en . Fo hyd ological modelling and lood o ecas ing, i is impo an o unde s and he spa ial- empo al a iabili y o soil mois u e a he basin le el (Co - adini, 2014; Kos e e al., 2010; Man eda e al., 2007; Ve- eecken e al., 2014). Physically based and dis ibu ed models can p o ide a lo o insigh in o how soil mois u e changes in space and ime as a unc ion o e ain, soil and ege a ion cha ac e is ics o he basin. Howe e , physically based dis ib- u ed models usually need a la ge numbe o pa ame e s (soil su ace, soil p ope ies and land use), which can inc ease he model unce ain y and dec ease he pe o mance o he model (Be en, 2006). The di ec g ound-based measu emen s o soil mois u e a e su icien ly accu a e, bu a e di icul , ime-consuming and limi ed o disc e e measu emen s a pa icula loca ions, which makes hem unsui able o hyd ological analyses a he basin le el (e.g. B occa e al., 2010; S i as a a e al., 2013; Wang and Qu, 2009). A use ul way o educe he unce ain y o he model and im- p o e he model pe o mance is o inco po a e emo ely sensed soil mois u e in o ma ion. Remo ely sensed soil mois u e p o- ides in o ma ion abou spa ial and empo al dynamics o soil mois u e, which can acili a e calib a ion and alida ion o hyd ological models o la ge scale (see e.g. Albe gel e al., 2010; B occa e al., 2011; Jackson e al., 2010; Pa ajka e al., 2006; Rö ze e al., 2014). Howe e , due o ela i ely coa se spa ial esolu ion o app oxima ely se e al ens o kilome e s, sa elli e soil mois u e obse a ions canno be e ec i ely ap- plied o hyd ological s udies in small ca chmen s. Due o a ia- ions in clima e, soil, ege a ion, opog aphy and o he ac o s, soil mois u e is he e ogeneously dis ibu ed wi hin ca chmen s. O e he pas decades, a ious downscaling me hods o sa - elli e soil mois u e p oduc s ha e been s udied o he im- p o emen o hei spa ial esolu ion. The a iabili y o soil mois u e wi hin a g id cell has o en been desc ibed by aking in o accoun soil ex u e (C ow e al., 2012; Gwak and Kim, 2017; Teuling and T och, 2005), ege a ion (Wes e n e al., 1999), opog aphy and o he impo an physical cha ac e is ics o he basin which e ec he soil mois u e (Hupe and Van- cloos e , 2002; Kos e e al., 2016; Rosenbaum e al., 2012). The es ima ion o sub- g id a iabil y o soil mois u e on he basis o soil ex u e da a is acili a ed due o he a ailabili y o high esolu ion da a on soil p ope ies o he en i e globe (Dai e al., 2019; Hengl e al., 2017; Shangguan e al., 2014; S oo ogel e al., 2017). A comp ehensi e e iew on he downscaling me hods o sa elli e emo e sensing based soil mois u e is p o ided in Peng Vesna Đukić, Ranka E ić, Mi osla Dumb o sky, Ve onika Sobo ko a 2 e al. (2017). They analysed he ad an ages and limi a ions associa ed wi h each me hod based on published alida ion s udies. Cu en ly, he e a e s ill no e ec i e ways o e alua - ing ei he he o iginal emo ely sensed soil mois u e o he downscaled soil mois u e ou pu s. Usually, he emo ely sensed soil mois u e p oduc s a e alida ed agains g ound–based soil mois u e obse a ions. In gene al, good ag eemen was ound be ween downscaled soil mois u e and in si u measu emen s (Lie ens e al., 2015; Peng e al., 2017; Ve hoes e al., 2015). I was also concluded ha he accu acy o he downscaled soil mois u e highly depends on he accu acy o he o iginal soil mois u e and ha i su passes he o iginal coa se soil mois u e o many s udies (Peng e al., 2017). Howe e , he consis ency and he le el o ag eemen be- ween downscaled soil mois u e and soil mois u e simula ed by dis ibu ed hyd ologic models, pa icual ly in small ca chmen s a e s ill no well unde s ood. The objec i e o his s udy is o e alua e he consis ency and he ag eemen be ween downscaled emo ely sensed soil mois u e, and soil mois u e es ima ed om he physically based and dis ibu ed SHETRAN hyd ological model in a small ca chmen . The e alua ion is pe o med o he small Jičinka Ri e ca chmen (75.9 km2) in he Czech Republic du ing lash loods. The e alua ion o modeled soil mois u e pa e ns and hei consis ency wi h sa el- li e da a du ing uno e en s p o ides addi ional in o ma ion abou model eliabili y and accu acy which is an essen ial s ep o he de elopmen o soil mois u e assimila ion s a egies in esea ch o ope a ional hyd ologic applica ions. 2 METHOD 2.1 SHETRAN model SHETRAN is a 3D coupled su ace/subsu ace physically based and spa ially dis ibu ed i e basin model. SHETRAN ( e sion V4.4.5) wa e low componen was used in his s udy. The wa e low componen consis s o 4 modules: e apo anspi- a ion/in e cep ion; o e land/channel; a iably sa u a ed subsu - ace and snowmel (Ewen e al., 2000). The componen s o in e cep ion and e apo anspi a ion we e neglec ed in his s udy, because hei in luence is negligible in ain e en mod- els. Bo h he o e land and channel lows a e desc ibed by he di usi e wa e app oxima ion o he ull S . Venan equa ions (Sain -Venan , 1871). The con inui y equa ion is as ollows: QA q x ∂∂ += ∂∂ (1) The momen um conse a ion is as ollows: 2 2 (( / )) 0 QQA hgQQ gA x x CAR α ∂∂ ∂ +++= ∂∂ ∂ (2) whe e: is ime, x is he dis ance measu ed along he channel (m), Q is discha ge (m3 s–1), A is he hyd aulic a ea (m2), q is he ibu a y ou low (m3 s–1), h is he channel dep h (m), C is he Chezy coe icien (m0.5 s–1), R is he hyd aulic adius (m) and α is he co ec ion ac o (–). The soil wa e mo emen in he unsa u a ed zone is de- sc ibed using he Richa ds equa ion (Richa ds, 1931). The A cGIS so wa e A cView 10.2 was used o p epa e he inpu da a ela ed o he physical cha ac e is ics o he basin and o he displaying and isualisa ion o he spa ially dis ibu ed soil mois u e alues ac oss he basin. 2.2 Calib a ion and alida ion o he SHETRAN model The hyd ological model was ob ained by he calib a ion and alida ion o he SHETRAN model on he basis o he measu ed s eam low hyd og aphs a he ca chmen ou le . The SHETRAN model was calib a ed o he s o m e en which happened in Sep embe 2007 and alida ed o he s o m e en s happened in June 2009, May 2010 and June 2010. In ain all– uno modelling, he s eam low is o c ucial impo ance because i e lec s he hyd ological esponse o he whole ca chmen . I was assumed ha he good ag eemen be ween he modeled and obse ed uno hyd og aphs implies ha he o he componen s o hyd ological cycle, in his case soil mois u e, a e app op ia ely de e mined. The a ea o ca chmen was disc e ized in o g id cells. The adop ed g id cell size in his s udy is 500 m × 500 m. Al hough i can be expec ed ha by adop ing a coa se g id esolu ion a lo o spa ially impo an da a may ha e been los , he use o a coa se g id esolu ion can be jus i ied when simula ing e en s o high in ensi y and/o hyd og aphs wi h a sho concen a ion ime (Molna and Julien, 2000), as i is he case in he analysed Jičinka Ri e ca chmen . In he s udy o Molna and Julien (2000), i was concluded ha a coa se g id esolu ion can be used in hyd ological models as long as pa ame e s a e app o- p ia ely calib a ed. Th ough nume ous simula ions i was concluded ha he S ickle ′s coe icien s o o e land low and o i e low, he e ical sa u a ed hyd aulic conduc i i y o he subsu ace soil and he sa u a ed wa e con en had an impo an e ec on he size o o e land low. I was also concluded ha he alues o base low we e dependan on he ho izon al sa u a ed hyd aulic conduc i i y in he sa u a ed zone (Đukić and Radić, 2014; Đukić and Radić, 2016). The p elimina y alues o model pa ame e s used in model calib a ion we e de e mined om he li e a u e (Table 3) and a e p esen ed in he Table 1. The alues o all o he model pa ame e s we e ixed and adop ed hei a e age alues om he li e a u e (Table 3 and Table 1). The adop ed alues o all pa ame e s a e p esen ed in Table 1. The ag eemen be ween he modelled and obse ed uno was e alua ed using ollowing objec i e unc ions. The i s objec i e unc ion is based on he o mula ion p oposed by Nash and Su cli e (1970) and is gi en by: CR1 = 1– 2 ,, 1 2 , 1 () () n obs i sim i i n obs i obs i QQ QQ = = − −   (3) whe e: Qobs,i is he obse ed s eam low on day i, Qsim,i is he simu- la ed s eam low, obs Qis he a e age o he obse ed s eam low o e he calib a ion (o e i ica ion) pe iods o n days. Due o changeable a iance o model e o s, he Nash – Su - cli e coe icien o e iciency ends o emphasize he la ge e o s. Fo compa ison, he unc ion o Chiew and McMahon (1994) was used as he second objec i e unc ion in which he squa e oo o he conside ed alues we e ela ed using he ollowing equa ion: () () 2 ,, 1 2 , 1 CR2 1 n obs i sim i i n obs i obs i QQ QQ = = − =− −   (4) Spa io- empo al analysis o emo ely sensed and hyd ological model soil mois u e in he small Jičinka Ri e ca chmen in Czech Republic 3 Table 1. The anges o model pa ame e s used in calib a ion o he SHETRAN model, i s op imal alues (in pa en hesis) and he adop ed alues o uncalib a ed pa ame e s o he Jičinka Ri e basin uncalib a ed pa ame e s o he Jičinka Ri e basin. Soil Type Soil/ ock pa ame e s Land use/ ege a ion O e land low/channel pa ame e s dep h (m) Tex u e k s (m day–1) khs (m day–1) θ s (–) θ (–) α (–) n (–) S (m1/3 s–1) S R (m1/3 s–1) Dys ic Cambisol 0–0.7 Sandy clay loam 0.223–0.5814 (0.300) 0.223–0.5814 (0.300) 0.419–0.695 (0.480) 0.047 0.014 1.317 Fo es 4–8 (7) 15–40 (30) 0.7–1.2 Sandy clay loam 0.223–0.5814 (0.300) 0.223–0.5814 (0.300) 0.419–0.695 (0.480) 0.047 0.014 1.317 A able land 8–20 (18) Geological subs a e 1.2–4 (sands one, schis s) 4 0.01–5 (4) 0.6 0.1 0.001 1.1 Na u al g asslands 7–18 (16) Rendzina 0–0.5 Clay loam 0.217–0.4105 (0.270) 0.217–0.4105 (0.270) 0.437–0.442 (0.440) 0.075 0.013 1.415 Sca ce ege a ion 30–50 (42) Geological subs a e 1.2–4 (sands one, schis s) 4 0.01–5 (4) Eu ic cambisol 0–0.5 Clay loam 0.217–0.4105 (0.270) 0.217–0.4105 (0.270) 0.426–0.469 (0.44) 0.075 0.013 1.415 0.5–1.05 Loam 0.128–0.192 (0.15) 0.15 0.426–0.469 (0.430) 0.078 0.036 1.56 Geological subs a e 1.2–4 (sands one, schis s) 4 0.01–5 (4) Flu isol 0–0.40 Clay loam 0.217–0.4105 (0.255) 0.217–0.4105 (0.255) 0.426–0.469 (0.430) 0.075 0.013 1.415 0.40–1.0 Sil y loam 0.130–0.196 (0.163) 0.163 0.452 0.093 0.005 1.68 1.0–1.25 Sandy clay loam 0.223–0.5814 (0.300) 0.223–0.5814 (0.300) 0.419–0.695 (0.480) 0.047 0.014 1.317 Geologic subs a e 1.25-4 (sands one, schis s) 4 0.01–5 (4) 0.6 0.1 0.001 1.1 The hi d in oduced c i e ion is po en ially use ul in he con ex o p edic ion, o example, whe e simula ions mus be as close as possible o he obse ed alues a each ime s ep (Ye e al., 1997). I is de ined by: ,, 1 , 1 CR3 1 n obs i sim i i n obs i obs i QQ QQ = = − =− −   (5) The ou h c i e ion (Pe ei a and P ui , 2004) quan i ies he abili y o he model o accu a ely ep oduce s eam low ol- umes o e he pe iods o obse a ion. C i e ion CR4 di e s om he o he h ee c i e ia (CR1–CR3), because i does no measu e de ia ion om he obse ed alues a each s ep o simula ion. The e o e, CR4 canno be used alone as a c i e ion o calib a ion. This c i e ion is de ined by: ,, 11 ,, 11 CR4 1 nn sim i obs i ii nn obs i sim i ii QQ QQ == ==    =− −      (6) In addi ion o he al eady men ioned ou c i e ia (CR1– CR4) he ollowing s a is ical measu es we e also used: he oo -mean-squa e e o (RMSE), he mean absolu e e o (MAE), he coe icien o co ela ion (R) and he index o ag eemen (d). They a e exp essed by he ollowing equa ions: MAE = ,, 1 1n obs i sim i i QQ n= −  (7) RMSE = 2 ,, 1 () n obs i sim i i QQ n = −  (8) R = ,, 1 22 ,, 11 ()() ()() n obs i obs sim i sim i nn obs i obs sim i sim ii QQQQ QQ QQ = == −− −−   (9) 2 1 2 1 () 1,01 () n obs sim i n sim obs obs obs i QQ dd QQ QQ = = − =− ≤ ≤ −+−   (10) 2.2 Downscaling app oach o sub-g id a iabili y es ima ion The sa elli e soil mois u e da a we e downscaled by applying he me hod de eloped by Qu e al. (2015) in his s udy. The downscaling app oach (Qu e al., 2015) applied in his s udy is based on he Mualem- an Genuch en (M G) model, in which he unsa u a ed soil hyd aulic p ope ies a e desc ibed using he model o an Genuch en (1980) o he wa e e en ion unc ion in combina ion wi h he hyd aulic conduc i i y unc ion in oduced by Mualem (1976). The soil wa e e en ion equa ion, θ (h), is gi en by: ( ) () 0 1 es m n S hh h θθ θθ α − =+ ≤  +  (11) whe e θ is he olume ic wa e con en (cm3 cm–3) a p essu e head h (cm); θ s and θ a e he esidual and sa u a ed wa e con- en (cm3 cm–3), espec i ely; α (cm–1), n (–), and m (–) (m = 1 – 1/n) a e shape pa ame e s. The hyd aulic conduc i i y unc- ion, K(h), is gi en by: 2 1/ () 1(1 ) , 0 Lmm ese e KS KS S h  =−− ≤  (12) whe e Ks is he sa u a ed hyd aulic conduc i i y (cm d–1) and K Vesna Đukić, Ranka E ić, Mi osla Dumb o sky, Ve onika Sobo ko a 4 (cm d –1) is he hyd aulic conduc i i y and L is he po e connec- i i y pa ame e (L = 0.5); Se is e ec i e sa u a ion gi en by: () e s h S θθ θθ − =− (13) Fo each g id o coa se scale sa elli e da a p oduc , he co e- la ion be ween he s anda d de ia ion and soil mois u e () θ σθ is exp essed as a unc ion o he mean and he s anda d de ia- ion o he soil hyd aulic pa ame e s (Ks, θ s, θ , α , n,) (Mon zka e al., 2018) using he ollowing equa ion: () ()() 2 222 22 2 13 12 22 22 22 2 02 23 22 22 1 34 12 23 22 111 22 11 n s nn n n n n aa bb aa aa aa b a a bb bb bb aa θ αα α α αα θ α σ σρ σρ σρ σρρρ σρ σρ σσ ρρ =   +++   +++          ++ + − + −    ++    (14) is he log- ans o med sa u a ed hyd aulic conduc i i y (ln Ks); ρ is he e ical co ela ion leng h o he espec i e pa ame e s. The coe icien s a1–a3 and b0–b4 a e ela ed o he mean o he soil hyd aulic pa ame e s: θ s, θ , h, α and n. They a e calcula ed using he equa ions which a e desc ibed in pape s (Mon zka e al., 2018; Qu e al., 2015). The sub-g id su ace soil mois u e alues can be es ima ed based on he known a e age alue o su ace soil mois u e wi hin a g id cell and using he es ima ed alue o he () θ σθ unc ion and p oxy in o ma ion. I is supposed ha he spa ial a iabili y o soil mois u e wi hin each coa se scale sa elli e pixel is ela ed o he known spa ial a iabili y o he p oxy da a (Mon zka e al., 2018). By mul iplica ion wi h he p o ided soil mois u e s anda d de ia ion a he gi en mean su ace soil mois u e, he sub-g id su ace soil mois u e alues can be cal- cula ed using he ollowing equa ion: () , , ˆ ij ij p PP θ θθσθσ − =+ (15) whe e: , ˆ ij θ is he p edic ed soil mois u e a his ine scale loca ion; Pi,j is he p oxy da a a he ine scale sub-g id y-loca ion i and x-loca ion j, Pis he mean o he p oxy, and σ P is he s anda d de ia ion o he p oxy. In his pape he su - ace soil mois u e da a we e downscaled om i s o iginal eso- lu ion o 25 km o 1 km esolu ion. This was done by using he sa u a ed hyd aulic conduc i i y as a p oxy o soil mois u e he e ogenei y. A e ha , he ob ained alues o soil mois u e we e esampled a a 500 m esolu ion. 2.4 Compa ison o su ace soil mois u e es ima es A e he calib a ion and alida ion o he SHETRAN model, simula ions o soil mois u e in unsa u a ed zone we e pe - o med using he se s o pa ame e s op imized o he calib a- ion and he alida ion ain e en s. In ha way, soil mois u e es ima es a e one o he esul s ob ained by applying he SHETRAN hyd ological model. The consis ency be ween he su ace soil mois u e simula ed by he hyd ologic model (SMHM) and he su ace soil mois u e downscaled om he sa elli e e ie ed soil mois u e p oduc (SMsca ) was spa ially analyzed using he same c i e ia which we e used o he e alua ion o he hyd ologic model pe o - mance. Howe e , in his case, ins ead o he simula ed (Qsim) and he obse ed s eam low alues (Qobs), he alues o SMHM and SMsca a e used in Equa ions (3) – (10). 3 DATA 3.1 S udy a ea The Jičinka Ri e basin up o he ″No y Jičin″ wa e le el moni o ing s a ion (Fig. 1) is si ua ed in he Mo a ian – Silesi- an Region, in he eas e n pa o he Czech Republic. The Jičin- ka Ri e is a ibu a y o he Mo a ice Ri e , which belongs o he basin o he Opa a Ri e and o he basin o he Bal ic Sea. Al i udes in he Jičinka Ri e basin a y om 270 me e s abo e sea le el in he lowe pa o he basin o 1000 me e s in he sou ce a eas o he basin. The s eep slopes o he e ain in he Jičinka Ri e basin wi h he a e age slope o 9.1% signi ican ly a ec he cha ac e is ics o uno . Fig. 1. The Jičinka Ri e basin wi h he i e sys em and he ain gauging and hyd ological s a ions in he basin. Fou pedological soil ypes iden i ied in he s udied basin in- clude he ollowing: lu isol (10.5%), eu ic cambisol (46.2%), dys ic cambisol (41.3%) and endzina (2%) (Fig. 2). The hy- d ogeological beha iou o he whole basin is de ined by he dominan p esence o e ia y and qua e na y ocks (sands ones, schis s, loess, sand and g a el) which occupy abou 75% o he basin (Fig. 2). Fou di e en ege a ion ypes iden i ied in he Jičinka Ri e basin a e o es (34%), na u al g asslands (48%), a able land (8.5%) and o cha ds (0.03%) (Fig. 2). U ban a eas occupy abou 9.4% o he basin a ea. Spa io- empo al analysis o emo ely sensed and hyd ological model soil mois u e in he small Jičinka Ri e ca chmen in Czech Republic 5 Fig. 2. The map o soil ypes, ege a ion ypes, geological ypes and he slope map o he Jičinka Ri e basin up o he "No y Jičin" wa e le el moni o ing s a ion. Table 2. The o al amoun s o p ecipi a ion (P o ) allen on he g ound, he olume o p ecipi a ion (Vp), he maximum uno alue (Qmax) and he uno olume (V ). Numbe o ain e en Simula ion pe iod P o (mm) Vp (103 m3) Qmax (m3/s) V (103 m3) 1 5.09.2007. (10:00) 17.09.2007. (10:00) 190.13 14429.9 98 6387.5 2 22.06.2009. (05:00 – 29.06.2009. (7:00) 150.2 11403 264 5805.5 3 11.05.2010. (13:00 – 29.05.2010. (3:00) 232.8 17670.4 75.6 15666.4 4 30.05.2010. (11:00 – 2.06.2010. (14:00) 46.2 3502.2 43.1 4429.9 3.2 Inpu da a o he SHETRAN model The a e age hou ly heigh s o ain all in he basin du ing he analyzed ain e en s we e calcula ed by applying he me hod o Thiessen polygons (Thiessen, 1911) based on he hou ly p ecip- i a ion le els measu ed a he clima ological s a ions in he basin (Hodsla ice, Ve o ice and No y Jičin) (Fig. 1). The hou ly alues o uno a e measu ed a he ″No y Jičin″ hyd o- logical s a ion. The cha ac e is ics o p ecipi a ion and uno a e p esen ed in Table 2. The soil, geological, ege a ion and slope maps o he Jičin- ka basin we e ob ained in he o m o ec o polygon da a a he co esponding digi ized maps (Fig. 2). All inpu da a and he co esponding inpu pa ame e s used in SHETRAN a e summa- ized in Table 3. Vesna Đukić, Ranka E ić, Mi osla Dumb o sky, Ve onika Sobo ko a 6 Table 3. Inpu da a and pa ame e s used in he SHETRAN model. 3.3 Sa elli e soil mois u e da a The sa elli e soil mois u e da a used in his s udy we e aken om he Eu opean Space Agency (ESA) Clima e Change Ini- ia i e (CCI) Soil Mois u e (SM) p ojec (h p://www.esasoilmois u e-cci.o g). The ESA CCI SM 04.7 p oduc consis s o h ee su ace soil mois u e da a se s: he “ACTIVE P oduc ”, he “PASSIVE P oduc ” and he “COM- BINED P oduc ” (Do igo e al., 2017). The ESA CCI SM p oduc was ob ained by combining he soil mois u e e ie als om se en passi e (SMMR, SSM/I, TMI, AMSR-E, WindSa , AMSR2 and SMOS) and wo ac i e (ERS AMI and ASCAT) mic owa e senso s in o a global da a se spanning he pe iod 1979 – 2010. The homogenized and me ged p oduc s p esen su ace soil mois u e wi h a global co e age and a coa se spa- ial esolu ion o app oxima ely 25 km and a high empo al esolu ion o 1 day (G ube e al., 2019). In his s udy he ACTIVE p oduc was used. The “ACTIVE P oduc ” was c ea ed by he Uni e si y o Vienna (TU Wien) based on obse a ions om C-band sca e ome e s. The senso s ERS AMI and ASCAT ope a e a simila equencies (5.3 GHz C-band) and sha e a simila design. Di e en algo i hms a e used o he e ie al o soil mois u e om sa elli e measu e- men s (G ube e al., 2017). The ACTIVE soil mois u e da a a e p o ided in e ms o sa u a ion deg ee [%], anging be ween 0 (d y) and 100 (sa u- a ed). In his s udy soil mois u e alues we e used in olume - ic uni s (m3/m3). Soil mois u e alues in olume ic uni s we e calcula ed by mul iplying he deg ee o sa u a ion by soil po os- i y (exp essed in m3/m3). Soil po osi y da a used in he con e - sion o olume ic soil mois u e measu emen s ha e been aken om he GLDAS-Noah da ase (Rodell e al., 2004). 3.4 Soil hyd aulic da a The soil hyd aulic p ope ies a e desc ibed using he ollowing pa ame e s: he sa u a ed wa e con en ( θ s), he esidual esidual wa e con en ( θ ), he sa u a ed hyd aulic conduc i i y (Ks), and he empi ical pa ame e s - anGenuch en - α and an Genuch en - n. These pa ame e s a e used in he SHETRAN hyd ological model, and hey a e also used in he applied p ocedu e o soil mois u e downscaling om sa elli e da a. The high esolu ion soil hyd aulic da a we e aken om he Global Soil Da ase o use in Ea h Sys em Models (GSDE) (Shangguan e al., 2014) loca ed a h p://globalchange.bnu.edu.cn/ esea ch/soil5.jsp. This da abase p o ides he global alues o soil hyd aulic and he mal pa am- e e s a he spa ial esolu ion o 30" and o e ical esolu ions o : 0 – 0.05 m, 0.05 – 0.15 m, 0.15 – 0.30 m, 0.30 – 0.60 m, 0.60 – 1.00 m, and 1.00 – 2.00 m. G id maps o hyd aulic soil pa ame e s o he Jičinka Ri e ca chmen , which we e de i ed om he GSDE da abase, a e p esen ed in Fig 3. 4 RESULTS AND DISCUSSION 4.1 Resul s o he hyd ologic model pe o mance e alua ion The compa a i e o e iew o he esul s o he calib a ion and alida ion o he SHETRAN model a e p esen ed in Fig. 4. The igu e shows a close ela ionship be ween he modeled and he obse ed uno hyd og aphs. The esul s o he assessmen o he quali y o model simula ions by applying CR1 – CR4 c i e ia in Eqs. (3) – (6), and by applying he s a is ical measu es in Eqs. (7) – (9) a e p esen ed in Table 4. On he basis o Table 4, i can be concluded ha he good ag eemen was ound be ween he obse ed and he modeled s eam low hyd og aphs o bo h he calib a ion and he alida- ion ain e en s. I should be no ed ha he bes pe o mance measu es we e obse ed o he calib a ion ain e en in Sep- embe 2007 acco ding o he majo i y o he applied c i e ia. A small d op in he de e mined alues o he applied c i e ia om calib a ion o e i ica ion indica es ha he model pe o mance dec eases only sligh ly. I should also be no ed ha he e was an imp o emen o he model pe o mance o he e i ica ion ain e en in June in 2010 compa ed o he calib a ion ain e en in Sep embe 2007, acco ding o he CR4 c i e ia, and acco ding o he ob ained alues o he Mean Absolu e E o (MAE) and he Roo Mean Squa e E o (RMSE). 4.2 Compa ison o soil mois u e es ima es Fo each o he analyzed ain e en s g id maps o daily al- ues o soil mois u e a a 500 m esolu ion we e c ea ed o he wo di e en soil mois u e sou ces. The spa ial pa e ns o he maximum su ace soil mois u e es ima ed o he days o he peak uno occu ing du ing he analysed ain e en s a e p e- sen ed in Fig. 5. On he basis o Fig 5. i can be seen ha he e is an ag ee- men be ween he soil mois u e alues simula ed by applica ion he SHETRAN model and he alues es ima ed by downscaling om he sca e ome e soil mois u e da a. As i can be ex- pec ed, he es ima ed spa ially dis ibu ed alues o soil mois- u e o he days o he peak uno occu ing du ing he ana- lysed ain e en s co espond o he alues o he sa u a ed wa e con en in majo i y o cases o he analysed ain e en s. Only in he case o he ain e en in June 2010, he a e age daily alues o es ima ed soil mois u e a e lowe in ega d o he alues o he sa u a ed wa e con en . Type o inpu da a Inpu pa ame e s Sou ce o inpu da a Me eo ological Hou ly p ecipi a ion Czech Hyd ome eo ological Ins i u e Hyd ological Regis e ed s eam low hyd og aphs Topog aphic Digi al ele a ion model (DEM) o he esolu ion: 10 m × 10 m T.G. Masa yk Wa e Resea ch Ins i u e Land use/ ege a ion dis ibu ion S ickle ’s coe icien o o e land low (S ) and o channel low (S R) (Engman,1986) Co ine Land Co e Da abases h ps://land.cope nicus.eu/pan-eu opean/co ine- land-co e Soil ypes Hyd aulic soil/ ock p ope ies (po osi y and speci ic s o age, esidual wa e con en ( θ ), sa u a ed wa e con en ( θ s), e i- cal sa u a ed hyd aulic conduc i i y (k s), ho izon al sa u a ed conduc i i y (khs), an Genuch en - α , an Genuch en - n h p://globalchange.bnu.edu.cn/ esea ch/soil5.jsp. Geological ype ″Czech Geological Su ey″-A cGIS online Spa io- empo al analysis o emo ely sensed and hyd ological model soil mois u e in he small Jičinka Ri e ca chmen in Czech Republic 7 Fig. 3. G id maps o hyd aulic soil pa ame e s o he Jičinka Ri e ca chmen de i ed om he GSDE da abase: sa u a ed wa e con en (cm3 cm–3), sa u a ed hyd aulic conduc i i y (cm day–1) esidual wa e con en (cm3 cm–3), and empi ical pa ame e s α (cm–1) and n (–). Fig. 4. Obse ed (Qobs) and he SHETRAN model simula ed (Qmod) s eam low hyd og aphs a he Jičinka Ri e moni o ing s a ion (″No y Jičin″ wa e le el moni o ing s a ion) o he calib a ion (Sep embe 2007) and alida ion (June 2009, May 2010 and June 2010) ain e en s. Table 4. The esul s o he applica ion o c i e ia CR1–CR4 and s a is ical measu es (MAE, RMSE, R and d) o he assessmen o he SHETRAN model simula ions. C i e ion Calib a ion ain e en Valida ion ain e en s Sep embe 2007 June 2009 May 2010 June 2010 CR1 0.940 0.905 0.810 0.872 CR2 0.895 0.747 0.803 0.759 CR3 0.720 0.533 0.602 0.577 CR4 0.897 0.786 0.908 0.926 MAE 2.082 3.550 3.412 1.348 RMSE 3.598 7.381 6.036 2.0349 R 0.975 0.961 0.908 0.965 d 0.986 0.978 0.951 0.973 0 2 4 6 8 10 12 0 20 40 60 80 100 120 5-Sep-2007 6-Sep-2007 7-Sep-2007 8-Sep-2007 9-Sep-2007 10-Sep-2007 11-Sep-2007 12-Sep-2007 13-Sep-2007 14-Sep-2007 15-Sep-2007 16-Sep-2007 17-Sep-2007 P Qobs Qsim P(mm) Q (m3/s) Da e Sep embe 2007 0 10 20 30 40 50 0 50 100 150 200 250 300 22-Jun-09 22-Jun-09 23-Jun-09 23-Jun-09 24-Jun-09 24-Jun-09 25-Jun-09 25-Jun-09 26-Jun-09 26-Jun-09 27-Jun-09 27-Jun-09 28-Jun-09 28-Jun-09 29-Jun-09 29-Jun-09 30-Jun-09 P(mm) Qsim Qobs Da e Q (m 3 /s) P(mm) June 2009 0 1 2 3 4 5 60 10 20 30 40 50 60 70 80 90 11-May-10 12-May-10 14-May-10 15-May-10 17-May-10 18-May-10 20-May-10 21-May-10 23-May-10 24-May-10 26-May-10 27-May-10 29-May-10 Ps (mm) Qobs Qsim May 2010 Q (m3/s P(mm) Da e 0 0.7 1.4 2.1 2.8 3.5 4.20 10 20 30 40 50 29-May-10 30-May-10 31-May-10 1-Jun-10 2-Jun-10 3-Jun-10 4-Jun-10 5-Jun-10 6-Jun-10 7-Jun-10 8-Jun-10 9-Jun-10 10-Jun-10 P Qobs Qsim June 2010 Q (m3/s) P(mm) Vesna Đukić, Ranka E ić, Mi osla Dumb o sky, Ve onika Sobo ko a 8 Fig. 5. The spa ial pa e ns o he maximum su ace soil mois u e simula ed by he SHETRAN hyd ologic model (le ) and es ima ed by downscaling om he sca e ome e ( igh ) o he days o he peak uno occu ing: 7 Sep embe 2007; 24 June 2009; 17 May 2010 and 2 June 2010. Table 5. The assessmen o he ag eemen be ween he wo spa ially dis ibu ed soil mois u e es ima es o he Jičinka Ri e ca chmen acco ding o c i e ion unc ions: CR3, CR4, MAE, R, RMSE and d (a e age/ minimum /maximum alues). C i e ion Calib a ion ain e en Valida ion ain e en s Sep embe 2007 June 2009 May 2010 June 2010 CR3 –0.28/–4.35/0.30 –0.495/–1.387/0 –1.45/–5.1/–0.02 –2.5/–7/–0.34 CR4 0.89/0.62/0.99 0.81/0.71/0.999 0.76/0.56/0.99 0.72/0.58/0.999 MAE 0.07/0.04/0.30.0 0.09/0.05/0.37 0.11/0.05/0.16 0.11/0.04/0.25 R 0.8/0.06/0.995 0.63/0.004/0.99 0.68/0.12/0.99 0.71/0.196/0.99 RMSE 0.096/0.08/0.196 0.11/0.06/0.15 0.12/0.05/0.18 0.13/0.05/0.17 d 0.55/0.18/0.66 0.51/0.15/0.91 0.49/0.24/0.68 0.41/0.15/0.53 Spa io- empo al analysis o emo ely sensed and hyd ological model soil mois u e in he small Jičinka Ri e ca chmen in Czech Republic 9 Fig. 6. The spa ial dis ibu ion o he co ela ion coe icien , he mean absolu e e o and he index o ag eemen ac oss he Jičinka Ri e ca chmen o he analyzed ain e en s in Sep embe 2007, June 2009, May 2010 and June 2010. The consis ency be ween he wo sou ces o soil mois u e in- o ma ion was spa ially analyzed in e ms o he same e alua ion c i e ia which we e used o he e alua ion o he hyd ological model pe o mance. The spa ial dis ibu ion o he co ela ion coe icien , he mean absolu e e o and he index o ag eemen a e shown in Fig. 6. The de e mined a e age, minimum and maximum alues o he applied e alua ion c i e ia a he le el o he Jičinka Ri e ca chmen a e also shown in Table 5. I should be no ed ha nega i e alues o c i e ia CR1 – CR3 we e ob ained when soil mois u e es ima es om he wo di - e en sou ces we e compa ed. The alues o c i e ia CR3, which anges om –7 o 0.3, a e shown in Table 5. Howe e ,