AIC and BIC for cosmological interacting scenarios

Fabiola Arevalo1∗, Antonella Cid1†and Jorge Moya3‡

1 Departamento de Ciencias Fısicas, Facultad de Ingenierıa y Ciencias,

Universidad de La Frontera, Temuco, Casilla 54-D, Chile

2 Departamento de Fısica and Grupo de Cosmologıa y Gravitacion GCG-UBB, Universidad del Bıo-Bıo,

Casilla 5-C, Concepcion, Chile and

3 Departamento de Fısica, Universidad de Concepcion, Casilla 160-C, Concepcion, Chile

August 24, 2017

Abstract

In this work we study linear and nonlinear cosmological interactions, which depend ondark matter and dark energy densities in the framework of General Relativity. By usingthe Akaike information criterion (AIC) and the Bayesian information criterion (BIC) withdata from SnIa (Union 2.1 and binned JLA), H(z), BAO and CMB we compare the inter-acting models among themselves and analyze whether more complex interacting models arefavored by these criteria. In this context, we find some suitable interactions that alleviatethe coincidence problem.

1 Introduction

Since the seminal work of Riess and Perlmutter [1], the astronomical observations of type Ia super-novae suggest that the late universe is in a phase of accelerated expansion driven by an unknowncomponent dubbed dark energy. The fundamental nature of this late accelerated expansion re-mains unexplained, nevertheless recent observations [2] are consistent with the simplest model,the ΛCDM scenario, which establishes that the energy density of the universe is dominated nowby a non-relativistic fluid (dark matter) and a cosmological constant (dark energy).

∗fabiola.arevalo@ufrontera.cl†acidm@ubiobio.cl‡jorgemoya@udec.cl

1

arX

iv:1

610.

0933

0v2

[as

tro-

ph.C

O]

22

Aug

201

7

Despite the observational success of the ΛCDM scenario, this model has theoretical problemssuch as the fine-tuning problem and the coincidence problem [3] also there are some observationaltensions recently reported, present when we use independently high redshift and low redshiftdata to constrain parameters [4]. Assuming that a departure of the ΛCDM scenario is needed, thesimplest generalization is the so-called ωCDM model, which describes dark energy as a perfect fluidwith a constant state parameter ω. Furthermore, models based on the interaction between darkmatter and dark energy have been studied to describe the accelerated expansion. One of the firstinteracting models was proposed in Ref.[5]; it was mainly motivated to alleviate the coincidenceproblem in an interacting-quintessence scenario, focusing in an asymptotic attractor behavior forthe ratio of the energy densities for the dark components. Since then, many interacting models withnumerical and analytical solutions have emerged [6]-[9], including interactions with change of signstudied in Refs.[10]-[12]. A detailed review of cosmological interactions can be found in Ref.[13]and some attempts to build an interaction from an action principle in Refs.[14]. In particularRefs.[15] present analytical solutions for a wide class of more elaborated interactions where thedark components are barotropic fluids with constant state parameters. Also, the question of howto discriminate among dark energy models (degeneracy problem [16]) has arisen in the context ofinteracting scenarios. In particular, there has been a debate on whether interacting models canbe distinguished from modified dark energy equations of state, Chaplygin gas or modified gravity[17], which remains an open issue.

To compare different models of a certain physical phenomenon in light of the data thereare criteria, based on Occam’s razor (“among competing hypotheses, the one with the fewestassumptions should be selected”). These criteria measure the goodness of fitted models comparedto a base model (see Refs.[18] and [19]). Two widely used criteria are the Akaike InformationCriterion (AIC) [20] and the Bayesian Information Criterion (BIC) [21]. The first is an essentiallyfrequentist criterion based on information theory and the second one follows from an approximationof the bayesian evidence valid for large sample size [18].

In Cosmology AIC and BIC have been applied to discriminate cosmological models basedon the penalization associated to the number of parameters that the model need to explain thedata. Specifically, in Ref.[22] the author performs cosmological model selection by using AIC andBIC in order to determinate the parameter set that better fit the WMAP3 data. Following thiswork in Ref.[23] the author considers more general models to the early universe description inlight of AIC and BIC, also including the deviance information criterion. Regarding late universedescription, the authors of Ref.[24] consider different models of dark energy and use informationcriteria to compare among them using the Gold sample of SnIa. Later on, the authors of [25]study interacting models, with an energy density ratio proportional to a power-law of the scalefactor attempting to alleviate the coincidence problem. By using AIC and BIC, they compare themodels among themselves and with ΛCDM considering data from SnIa, BAO and CMB. Morerecently, in Ref.[26] the authors find that a particular interacting scenario is disfavored compared

2

to ΛCDM. They study an interaction proportional to a power-law of the scale factor, by usingAIC and BIC, and considering data from SnIa, H(z), BAO, Alcock–Paczynski test and CMB.

In this work we analyze eight general types of interacting models with analytical solution usingUnion 2.1 (or binned JLA)+H(z)+BAO+CMB data under AIC and BIC. The main goal of ourwork is to investigate if complex interacting models are competitive in fitting the data and whetherwe could distinguish among them via the model comparison approach.

This paper is organized as follows: in section 2 we present and motivate eight types of inter-acting models with analytical solution to be revised. In section 3 we show the functions to befitted and describe the information criteria to be used. In section 4 we present the analysis andresults of the data fitting process and finally in section 5 we discuss our final remarks.

2 Interacting Models

We work in the framework of general relativity by considering a spatially flat Friedmann-Lemaıtre-Robertson-Walker universe. The Friedmann equation is written as

3H2 = ρ, (1)

where H = a/a is the Hubble expansion rate, a is the scale factor, the dot represents a derivativewith respect to the cosmic time and we have considered 8πG = c = 1. From the energy-momentumtensor conservation we have

ρ+ 3H(ρ+ p) = 0, (2)

where ρ is the total energy density and p is the effective pressure. First we consider that darkmatter and dark energy are the relevant components of the total energy density at late times,i.e., ρ = ρx + ρm and p = px + pm (where the subscripts x and m represent dark energy (DE)and dark matter (DM), respectively). Furthermore, we consider a barotropic equation of state forboth fluids, i.e., px = ωxρx and pm = ωmρm. To include a phenomenological interaction betweenthese fluids, we separate the conservation Eq.(2) into two equations

ρm + 3γmHρm = −Q, (3)

ρx + 3γxHρx = Q, (4)

where γx = 1 + ωx, γm = 1 + ωm and Q represents the interaction function between dark matterand dark energy. Using the change of variable η = 3 ln a and defining ()′ := d/dη, Eqs.(3) and (4)are rewritten as

ρ′m + γmρm = −Γ, (5)

ρ′x + γxρx = Γ, (6)

3

with Γ = Q/3H. For Γ > 0 we have an energy transfer from DM to DE and for Γ < 0 we havethe opposite energy transfer, from DE to DM. From Eqs.(5) and (6) and considering ρ = ρx + ρm

we can write ρx and ρm as [15]:

ρx =γmρ+ ρ′

∆, ρm = −γxρ+ ρ′

∆, (7)

with ∆ = γm − γx and from Eq.(2) we get that

p = −ρ− ρ′. (8)

From Eqs.(5) and (7) we obtain the “source equation” defined in Ref.[15]:

ρ′′ + (γx + γm)ρ′ + γxγmρ = ∆Γ, (9)

valid for γx and γm constants. We notice that due to (7) every Γ proportional to ρx and/or ρm

in (9) constitutes in fact, a differential equation for the variable ρ. Also, it is worth to mentionthat Eq.(9) can be rewritten as a differential equation in terms of the deceleration parameter orin terms of a variable state parameter in a holographic context [27].

In this work we study eight types of interaction [10, 15], defined as: Γ1 = αρm + βρx, Γ2 =αρ′m + βρ′x, Γ3 = αρmρx/(ρm + ρx), Γ4 = αρ2

m/(ρm + ρx), Γ5 = αρ2x/(ρm + ρx), Γ6 = αρ, Γ7 = αρ′

and Γ8 = αqρ = −α(ρ + 3ρ′/2), where q = −(

1 + HH2

)is the deceleration parameter, ρ is the

total energy density and α, β are constants.By rewriting Eq.(9) as

ρ [ρ′′ + b1ρ′ + b3ρ] + b2ρ

′2 = 0, (10)

it includes the eight types of interaction we are interested in, where the constants b1, b2, b3 aredifferent combinations of the relevant parameters depending on the particular interaction; seeTable 1. The general solution of Eq.(10) takes the form

ρ(a) =[C1a

3λ1 + C2a3λ2] 1

1+b2 . (11)

The integration constants in (11) are given by

C1 = −(3H20 )1+b2

[λ2 + γ0(1 + b2)

λ1 − λ2

],

C2 = (3H20 )1+b2

[λ1 + γ0(1 + b2)

λ1 − λ2

], (12)

4

and

λ1 = −1

2

(b1 +

√b2

1 − 4b3(1 + b2)

),

λ2 = −1

2

(b1 −

√b2

1 − 4b3(1 + b2)

),

γ0 = γm − Ωx0∆, (13)

where H0 and Ωx0 are the Hubble parameter and the value of the density parameter for DE today(i.e. Ωx0 = ρx0/3H

20 ), respectively.

Table 1: Definition of the constants b1, b2 and b3 in terms of the relevant parameters for thestudied interactions.

Interaction b1 b2 b3

Γ1 = αρm + βρx γm + γx + α− β 0 γmγx + αγx − βγm

Γ2 = αρ′m + βρ′xγm + γx + αγx − βγm

1 + α− β0

γmγx

1 + α− βΓ3 = αρmρx/(ρm + ρx) γm + γx + α

γm + γx

∆

α

∆γmγx + α

γmγx

∆

Γ4 = αρ2m/(ρm + ρx) γm + γx −

2αγx

∆− α

∆γmγx −

αγ2x

∆

Γ5 = αρ2x/(ρm + ρx) γm + γx −

2αγm

∆− α

∆γmγx −

αγ2m

∆Γ6 = αρ γm + γx 0 γmγx − α∆

Γ7 = αρ′ γm + γx − α∆ 0 γmγx

Γ8 = αqρ = −α(ρ+ 32ρ′) γm + γx +

3

2α∆ 0 γmγx + α∆

The nature of cosmic interaction remains unknown, however, physical motivation to study mostof the interactions in Table 1 can be found in the literature. These interactions are worth to studybecause it has been shown that most of them could alleviate the coincidence problem [15, 28]. Itwas demonstrated in Ref.[29] that an interaction proportional to Hρx could be consistent withthe second law of thermodynamics if the energy transfer is from DE to DM, also, in Ref.[30] itwas shown that interactions proportional to H(ρm + ρx) or Hρm can arise by imposing simplethermodynamic arguments based on the evolution of the ratio ρm/ρx. For interactions proportionalto ρ′m, ρ′x or a linear combination of both, we note from Eqs.(3) and (4), that these interactions

5

can be rewritten in terms of interactions proportional to a linear combination of ρm and ρx. Wecan find a physical motivation to nonlinear interactions in Ref.[31], in the context of holographicinteracting models. On the other hand, sign-changeable interaction was found to be preferredby the data in Refs.[11]-[12]. It has also been shown that a late-time interaction can alleviatethe tension that arises in ΛCDM between the Hubble constant measurements from Planck andthe Hubble Space Telescope [32]. In Refs.[33] it was shown that interaction proportional to Hρm,Hρx and Hρmρx/(ρm+ρx) can have stable cosmological perturbations during the whole expansionhistory, i.e. these interactions could consistently describe the linear evolution of growing structures,without large-scale instabilities.

On the other hand, the effective energy density (11) associated to the general solution of ourinteractions has an effective pressure (8) corresponding to a variable modified Chaplygin gas [34]given by

p = −ρ(

1 +λ1

1 + b2

)− C2

λ2 − λ1

1 + b2

ρ−b2a3λ2 . (14)

This means that the considered interactions can be interpreted as a single fluid model in aunified description of the dark sector inherently.

Also, the effective energy density (11) can be interpreted as a non-interacting description ofthe dark sector with a variable barotropic index for the dark energy component given by

γx(a) = −C1λ1a3λ1 + C2λ2a

3λ2 + γmρm0a−3γm

[C1a3λ1 + C2a3λ2 − ρm0a−3γm + ρx0], (15)

where ρm0 and ρx0 are, respectively, the current values of the DM and DE densities. The in-verse approach has been considered in Ref.[35], where the relation between a given variable stateparameter and a reconstructed interaction has been addressed using Gaussian processes.

The solution in Eq.(11) is valid for late-time evolution, nevertheless if we are interested in datafrom BAO and/or CMB, which consider high redshifts, we need to take into account the radiationcontribution in the equations as well as the baryons contribution. If we consider from here onρ = ρm + ρx + ρr + ρb, with ρr the energy density of relativistic matter and ρb the energy densityof baryons, which we assume are non-interacting with the dark fluids, then the solution of Eq.(10)is given by

ρ(a) =[C1a

3λ1 + C2a3λ2] 1

1+b2 + 3H20

(Ωr0

a4+ Ωb0

a3

), (16)

where Ωr0 and Ωb0 are the current values of the density parameters for radiation and baryons,respectively, and the constants C1 and C2 (for interactions Γ1 to Γ5) are modified to

C1 =[3H2

0 (Ωx0 + Ωm0)]1+b2 − C2,

6

C2 = −(3H20 )1+b2 [(Ωx0γx + Ωm0γm)(1 + b2)]

(Ωx0 + Ωm0)−b2(λ2 − λ1)− (3H2

0 )1+b2λ1

(λ2 − λ1)(Ωx0 + Ωm0)−1−b2. (17)

The values of b1, b2, b3 are the same for both cases, including radiation and baryons or not; seeTable 1.

For interactions Γ6−Γ8 we can decompose the general solution into a homogeneous solution ρh

and a particular solution ρp, then the general solution is given by ρ = ρh + ρp. The homogeneouspart of the solution ρh corresponds to (16) and the particular solution is given by

ρp(a) = −9Mria−4 −Mbia

−3, (18)

where Mri = −3H20δriΩr0∆/(12b1 − 9b3 − 16), Mbi = 3H2

0δbiΩb0∆/(2b1 − 2b3 − 2), (δr6, δr7, δr8) =(−α, 4

3α,−α

), (δb6, δb7, δb8) = (2α,−2α, α) and now the constants C1 and C2 are given by

C1 = 3H20 (Ωx0 + Ωm0) + 9Mri +Mbi − C2, (19)

C2 =3H2

0 ∆Ωx0

λ2 − λ1

− (9λ1 + 12)Mri

λ2 − λ1

− (λ1 + 1)Mbi

λ2 − λ1

− 3H20 (Ωx0 + Ωm0)(γm + λ1)

λ2 − λ1

. (20)

Additionally, to examine the coincidence problem we use the coincidence parameter r definedas

r =ρm

ρx

. (21)

We can therefore calculate the asymptotic limit of r(a) when a tends to∞. For all our interactionswe get

r∞ = −

[1 +

2(γx − 1)(1 + b2)

2(1 + b2)− b1 +√b2

1 − 4b3(1 + b2)

], (22)

a constant that depends on the state parameters and interaction parameters. The author ofRef.[15] noticed that, for a constant and positive γx and for an interacting term proportional to ρ,ρ′ or ρx, there is obtained a positive r parameter asymptotically constant, alleviating in this sensethe coincidence problem. Furthermore, the authors in Ref. [28], analyze nonlinear models Γ3, Γ4

and Γ5, concluding that the last two interactions may alleviate the coincidence problem also.In this section we have assumed that an interacting scenario of DM and DE can be described

in terms of fluids with a constant state parameter. In this sense, the source equation (9) allowsus to study a family of interacting scenarios recast in a single functional form (11), where wehave considered the more common linear and nonlinear interactions and also a naturally sign-changeable interaction. Besides, these interactions can be interpreted, at the background level, interms of a unified fluid description with a variable modified Chaplygin gas (14) or, in terms of avariable equation of state (15) for the dark energy component with a non-interacting dark sector.

7

3 Observational analysis and model selection

In order to constrain the interacting models, we use the following data: i) distance modulus oftype Ia supernovae from: 580 data points from the Union 2.1 compilation [36] or 31 data points ofbinned data from the JLA compilation [37], ii) 28 data points from H(z) data [38]. iii) For BAOdata we use: the acoustic parameter (3 data points from the WiggleZ experiment [39]) and thedistance ratio (2 data points from the SDSS [40] and 1 data point from the 6dFGS surveys [41]).From CMB data we consider the position of the first peak in the CMB anisotropy spectrum [42].

To fit the cosmological models to the data we use the Chi-square method. Each dataset (SnIa,H(z), WiggleZ, SDSS, 6dFGS and CMB) has a corresponding Chi-square function (χ2

Sn, χ2H(z),

χ2WiggleZ, χ2

SDSS, χ26dFGS, χ2

CMB) which is used to calculate the overall χ2 function. These functionsare defined according to each dataset.

For SnIa we have the χ2 function defined as

χ2Sn =

NSn∑i=1

(µi,th − µi,obs)2

σ2µi

, (23)

where µ is the distance modulus defined in appendix (A1), “th” represents the theoretical function,“obs” the observed value, σµi is the uncertainty associated to the observed value and NSn is thedata number of SnIa in the compilation of Union 2.1 or the number of binned data for the JLAcompilation. Similarly, for H(z) we have the χ2 function for the Hubble expansion rate (A3):

χ2H(z) =

NH∑i=1

(Hi,th −Hi,obs)2

σ2Hi

, (24)

where NH is the data number of H(z) data.For BAO’s measurements we have χ2

BAO given by

χ2BAO = χ2

WiggleZ + χ2SDSS + χ2

6dFGS. (25)

In the case of WiggleZ we use the inverse of the covariance matrix C−1WiggleZ [39],

χ2WiggleZ = (Ath − Aobs)C

−1WiggleZ(Ath − Aobs)

T , (26)

where Ath is the theoretical acoustic parameter defined in the appendix (A4), the observationalvalues of this parameter are given by Aobs = (0.474, 0.442, 0.424) at redshifts z = (0.44, 0.6, 0.73),respectively, and

C−1WiggleZ =

1040.3 −807.5 336.8−807.5 3720.3 −1551.9336.8 −1551.9 2914.9

. (27)

8

Analogously, for SDSS [40] we have

χ2SDSS = (dth − dobs)C

−1SDSS(dth − dobs)

T , (28)

where dth is the theoretical distance ratio defined in the appendix, see Eq.(A7), the observationalvalues are given by dobs = (0.1905, 0.1097) at redshifts z = (0.2, 0.35) and the inverse of thecovariance matrix is

C−1SDSS =

(30124 −17227−17227 86977

). (29)

The data point of the 6dFGS is given by

χ26dFGS =

(dth − dobs

σd

)2

, (30)

with the observed distance ratio dobs = 0.336 and σd = 0.015, at redshift z = 0.106 [41].Finally, we consider the position of the first peak of the CMB anisotropy as a background data

coming from early universe’s physics. It is common to consider also the shift parameter, but thederivation of this parameter is assuming a ΛCDM scenario today [43]. It is more consistent toconsider only the position of the first peak to test interacting models because it only depends onpre-recombination physics (see the discussion in Refs. [44]) and in this sense, it can be consideredin our work as a good approximation. The χ2 contribution of the position of the first peak l1 isgiven by

χ2CMB =

(l1th − l1obs

σl

)2

, (31)

where l1th is the position of the first peak defined in the appendix (A11), l1obs is the observedposition of the first peak, l1obs = 220.0 and σl = 0.5 [42].

In order to find the best fit model parameters we perform a joint analysis using all the data,we minimize the overall χ2 function defined as

χ2 = χ2Sn + χ2

H(z) + χ2BAO + χ2

CMB. (32)

Each Chi-squared function depends on the parameters of the model. Based on statisticalanalysis we can determine which models are “better” taking into account how many parametersdo the models need and how well do they fit the data. In this work we use two criteria, the AkaikeInformation Criterion (AIC) and the Bayesian Information Criterion (BIC). The AIC parameteris defined through the relation [20]:

AIC = χ2min + 2d, (33)

9

where d is the number of free parameters in the model and χ2min is the minimum value of the χ2

function. The “preferred model” for this criterion is the one with the smaller value of AIC. Thiscriterion “penalizes” models according to the number of free parameters that they have.

To compare the model k with the model l, we calculate ∆AICkl = AICk −AICl, which can beinterpreted as “evidence in favor” of the model k compared to the model l. For 0 ≤ ∆AICkl < 2we have “strong evidence in favor” of model k, for 4 < ∆AICkl ≤ 7 there is “little evidence infavor” of the model k, and for ∆AICkl > 10 there is basically “no evidence in favor” of model k[23].

On the other hand, the Bayesian criterion is defined through the relation

BIC = χ2min + d lnN, (34)

where N is the number of data points. Similarly to ∆AICkl, ∆BICij = BICi − BICj can beinterpreted as “evidence against” the model i compared to the model j. For 0 ≤ ∆BICij < 2there is “not enough evidence against” the model i, for 2 ≤ ∆BICij < 6 there is “evidence against”the model i and for 6 ≤ ∆BICij < 10 there is “strong evidence against” model i [23].

4 Analysis and results

For model fitting we use the Chi-Square method with the Levenberg–Marquardt algorithm im-plemented in the package lmfit of Python.1 For all the studied interactions we consider a fixedγm. The search ranges of the free parameters in our models are: Ωm ∈ [0, 1], γx ∈ [−0.5, 0.5],α ∈ [−0.5, 0.5], β ∈ [−0.5, 0.5] and h ∈ [0, 1]. We use the combined datasets Union 2.1 (or binnedJLA), H(z), BAO and CMB for the data fitting and we restrict our analysis to a maximum of fourfree parameters for each model.

We consider two possible scenarios, one where we fix parameters such as γm = 1 which corre-sponds to a cold dark matter scenario or we fix γm = 1 and γx = 0 that corresponds to a Λ(t)CDMmodel [45]. For these scenarios we can additionally fix the parameters associated with differentmodels of phenomenological interaction, α and/or β.

In Table 2 the best fit parameters for all the analyzed models are shown; we used a jointanalysis considering Union 2.1+H(z)+BAO+CMB. The subscripts a, b, c, d, e, f , g in the modelsdenote γx = 0, α = 0, β = 0, α = β, γx = α = 0, γx = β = 0 and γx = 0 with α = β, respectively.From Table 1 and in the context of this classification we note that Γ2e does not correspond to aninteracting model, because the parameters b1, b2 and b3 in Table 1 have fixed values in this case.Because of this, Γ2e is not present in Tables 2 - 5. Also, we note that the only difference betweenΓ1f and Γ2g is a sign in the interaction term, thus we exclude Γ2g from the analysis.

1https://www.python.org

10

Tab

le2:

Res

ult

sof

the

dat

afitt

ing

usi

ng

the

join

tan

alysi

sfr

omU

nio

n2.

1,H

(z),

BA

Oan

dC

MB

.T

he

erro

rin

form

edco

rres

pon

ds

to68

%co

nfiden

cele

vel.

Fix

edm

eans

that

the

par

amet

erw

asse

tto

zero

and

the

das

hed

lines

mea

nth

atth

em

odel

does

not

hav

eth

atpar

amet

er.

The

der

ived

par

amet

ers

are:

the

curr

ent

valu

eof

the

dec

eler

atio

npar

amet

erq 0

,th

eva

lue

ofth

eeff

ecti

vest

ate

par

amet

erto

day

weff

and

the

calc

ula

ted

age

ofth

euniv

erse

inG

y.T

he

AIC

and

BIC

par

amet

ers

are

indic

ated

inea

chca

se.

Mod

elΩ

m0

γx

αβ

hq 0

ωeff

Age

AIC

BIC

Γ1a

0.23

9±

0.02

1F

ixed

0.00

04±

0.00

280.

0060±

0.04

230.

699±

0.00

3−

0.57

3±

0.03

1−

0.71

6±

0.02

113.6

33±

0.41

058

8.45

060

6.13

7

Γ1b

0.24

7±

0.02

7−

0.05

9±

0.09

9F

ixed

0.00

45±

0.00

440.

701±

0.00

4−

0.62

4±

0.11

4−

0.74

9±

0.07

613.6

16±

0.42

058

7.42

260

5.10

8

Γ1c

0.25

0±

0.01

6−

0.06

1±

0.05

80.

0000±

0.00

49F

ixed

0.70

1±

0.00

4−

0.62

2±

0.06

6−

0.74

8±

0.04

413.6

18±

0.24

958

7.43

660

5.12

3

Γ1d

0.25

0±

0.02

2−

0.06

0±

0.08

60.

0000±

0.00

200.

0000±

0.00

200.

701±

0.00

4−

0.62

1±

0.09

8−

0.74

7±

0.06

513.6

18±

0.34

758

7.43

560

5.12

2

Γ1e

0.23

9±

0.01

4F

ixed

Fix

ed0.

0010±

0.00

060.

699±

0.00

3−

0.57

3±

0.02

1−

0.71

5±

0.01

413.6

72±

0.19

158

6.50

259

9.76

7

Γ1f

0.24

1±

0.01

4F

ixed

0.00

03±

0.00

01F

ixed

0.69

9±

0.00

3−

0.57

0±

0.02

1−

0.71

3±

0.01

413.6

50±

0.19

058

6.46

459

9.72

9

Γ1g

0.24

1±

0.01

3F

ixed

0.00

03±

0.00

150.

0003±

0.00

150.

699±

0.00

3−

0.57

0±

0.01

9−

0.71

3±

0.01

313.6

47±

0.17

458

6.46

259

9.72

6

Γ2a

0.23

8±

0.01

5F

ixed

0.00

04±

0.00

53−

0.24

02±

52.2

394

0.69

9±

0.00

3−

0.57

5±

0.02

2−

0.71

7±

0.01

513.6

95±

0.23

258

8.66

660

6.35

2

Γ2b

0.24

9±

0.03

1−

0.05

8±

0.06

7F

ixed

0.00

33±

0.89

060.

701±

0.00

4−

0.62

0±

0.08

6−

0.74

7±

0.05

813.6

20±

0.57

058

7.43

460

5.12

0

Γ2c

0.25

1±

0.01

9−

0.06

4±

0.07

1−

0.00

01±

0.00

26F

ixed

0.70

1±

0.00

4−

0.62

4±

0.08

1−

0.74

9±

0.05

413.6

05±

0.29

358

7.45

660

5.14

3

Γ2d

0.24

7±

0.02

0−

0.05

4±

0.07

30.

0003±

0.00

380.

0003±

0.00

380.

701±

0.00

4−

0.61

9±

0.08

4−

0.74

6±

0.05

613.6

48±

0.31

458

7.48

960

5.17

5

Γ2f

0.24

2±

0.01

4F

ixed

−0.

0003±

0.00

01F

ixed

0.69

9±

0.00

3−

0.56

9±

0.02

1−

0.71

3±

0.01

413.6

42±

0.19

058

6.46

759

9.73

2

Γ3

0.25

1±

0.02

4−

0.03

0±

0.07

70.

0005±

0.00

17−−−

0.70

0±

0.00

4−

0.58

7±

0.09

0−

0.72

5±

0.06

013.5

61±

0.35

958

8.32

260

6.00

9

Γ3a

0.24

5±

0.01

4F

ixed

0.00

04±

0.00

01−−−

0.69

8±

0.00

3−

0.56

4±

0.02

1−

0.71

0±

0.01

413.6

02±

0.19

158

6.62

559

9.89

0

Γ4

0.25

4±

0.01

7−

0.06

8±

0.06

70.

0005±

0.00

23−−−

0.70

1±

0.00

4−

0.62

2±

0.07

6−

0.74

8±

0.05

013.5

68±

0.26

958

7.69

260

5.37

9

Γ4a

0.24

0±

0.01

4F

ixed

0.00

01±

0.00

01−−−

0.69

9±

0.00

3−

0.57

1±

0.02

1−

0.71

4±

0.01

413.6

62±

0.18

958

6.47

659

9.74

1

Γ5

0.25

0±

0.02

3−

0.05

9±

0.08

4−

0.00

40±

0.00

38−−−

0.70

1±

0.00

4−

0.61

9±

0.09

6−

0.74

6±

0.06

413.6

22±

0.35

758

7.44

960

5.13

5

Γ5a

0.23

5±

0.01

3F

ixed

0.01

58±

0.00

49−−−

0.69

9±

0.00

3−

0.58

0±

0.01

9−

0.72

0±

0.01

313.6

64±

0.17

758

6.62

059

9.88

5

Γ6

0.24

3±

0.01

8−

0.04

4±

0.06

60.

0016±

0.00

17−−−

0.70

1±

0.00

4−

0.61

5±

0.07

6−

0.74

3±

0.05

113.6

47±

0.29

058

7.44

060

5.12

6

Γ6a

0.23

6±

0.01

4F

ixed

0.00

19±

0.00

09−−−

0.69

9±

0.00

3−

0.57

7±

0.02

1−

0.71

8±

0.01

413.6

72±

0.19

658

6.11

959

9.38

4

Γ7

0.24

4±

0.01

9−

0.04

6±

0.04

6−

0.00

16±

0.00

79−−−

0.70

1±

0.00

3−

0.61

6±

0.05

8−

0.74

4±

0.03

913.6

51±

0.29

258

7.42

160

5.10

7

Γ7a

0.23

7±

0.01

3F

ixed

−0.

0018±

0.00

06−−−

0.69

9±

0.00

3−

0.57

6±

0.02

0−

0.71

7±

0.01

313.6

77±

0.18

458

6.10

359

9.36

7

Γ8

0.23

0±

0.01

8−

0.01

8±

0.06

30.

0012±

0.00

21−−−

0.70

1±

0.00

4−

0.60

6±

0.07

4−

0.73

8±

0.04

913.8

05±

0.29

958

9.30

560

6.99

1

Γ8a

0.23

9±

0.01

4F

ixed

0.00

34±

0.00

13−−−

0.69

9±

0.00

3−

0.57

3±

0.02

0−

0.71

5±

0.01

413.6

79±

0.19

058

6.09

759

9.36

2

ωC

DM

0.24

9±

0.01

6−

0.05

9±

0.08

1−−−

−−−

0.70

1±

0.00

4−

0.62

1±

0.09

0−

0.74

7±

0.06

013.6

20±

0.27

458

5.43

559

8.70

0

ΛC

DM

0.23

9±

0.00

7−−−

−−−

−−−

0.69

9±

0.00

3−

0.57

2±

0.01

0−

0.71

5±

0.00

713.6

73±

0.10

058

4.50

559

3.34

8

11

Tab

le3:

Res

ult

sof

the

dat

afitt

ing

usi

ng

the

join

tan

alysi

sfr

omU

nio

n2.

1,H

(z)

and

BA

O.

The

erro

rin

form

edco

rres

pon

ds

to68

%co

nfiden

cele

vel.

Fix

edm

eans

that

the

par

amet

erw

asse

tto

zero

and

the

das

hed

lines

mea

nth

atth

em

odel

does

not

hav

eth

atpar

amet

er.

The

der

ived

par

amet

ers

are:

the

curr

ent

valu

eof

the

dec

eler

atio

npar

amet

erq 0

,th

eva

lue

ofth

eeff

ecti

vest

ate

par

amet

erto

dayw

effan

dth

eca

lcula

ted

age

ofth

euniv

erse

inG

y.T

he

AIC

and

BIC

par

amet

ers

are

indic

ated

inea

chca

se.

Mod

elΩ

m0

γx

αβ

hq 0

ωeff

Age

AIC

BIC

Γ1a

0.24

3±

0.02

6F

ixed

0.00

64±

0.01

12−

0.03

00±

0.06

990.

699±

0.00

3−

0.56

7±

0.04

0−

0.71

1±

0.02

613.7

87±

0.68

758

7.79

060

5.47

0

Γ1b

0.24

7±

1.60

8−

0.05

9±

2.37

7F

ixed

0.00

49±

2.43

620.

701±

0.00

4−

0.62

4±

3.59

1−

0.75

0±

2.39

413.6

14±

27.3

4858

7.42

260

5.10

2

Γ1c

0.24

6±

0.04

4−

0.05

3±

0.13

40.

0012±

0.01

28F

ixed

0.70

1±

0.00

4−

0.61

8±

0.15

8−

0.74

6±

0.10

513.6

37±

0.65

558

7.39

860

5.07

8

Γ1d

0.24

6±

0.04

9−

0.05

3±

0.14

10.

0010±

0.01

270.

0010±

0.01

270.

701±

0.00

4−

0.61

9±

0.16

9−

0.74

6±

0.11

213.6

33±

0.73

458

7.40

060

5.08

0

Γ1e

0.23

8±

0.02

2F

ixed

Fix

ed0.

0045±

0.05

190.

699±

0.00

3−

0.57

4±

0.03

3−

0.71

6±

0.02

213.6

54±

0.47

658

6.49

059

9.75

0

Γ1f

0.23

5±

0.01

5F

ixed

0.00

37±

0.00

94F

ixed

0.69

9±

0.00

3−

0.57

9±

0.02

2−

0.71

9±

0.01

513.6

87±

0.21

958

6.04

859

9.30

8

Γ1g

0.23

5±

0.01

6F

ixed

0.00

32±

0.00

860.

0032±

0.00

860.

699±

0.00

3−

0.58

0±

0.02

4−

0.72

0±

0.01

613.6

76±

0.23

758

6.10

859

9.36

8

Γ2b

0.24

7±

1.69

5−

0.05

9±

2.50

5F

ixed

0.07

35±

40.0

856

0.70

1±

0.00

4−

0.62

4±

3.78

4−

0.74

9±

2.52

313.6

15±

29.4

2658

7.42

260

5.10

2

Γ2c

0.24

6±

0.04

4−

0.05

3±

0.13

4−

0.00

12±

0.01

28F

ixed

0.70

1±

0.00

4−

0.61

9±

0.15

8−

0.74

6±

0.10

513.6

36±

0.65

458

7.39

860

5.07

8

Γ2d

0.24

6±

0.04

3−

0.05

2±

0.13

4−

0.00

12±

0.01

28−

0.00

12±

0.01

280.

701±

0.00

4−

0.61

8±

0.15

8−

0.74

6±

0.10

513.6

37±

0.65

458

7.39

860

5.07

8

Γ2f

0.23

5±

0.01

5F

ixed

−0.

0037±

0.00

94F

ixed

0.69

9±

0.00

3−

0.57

9±

0.02

2−

0.71

9±

0.01

513.6

87±

0.21

958

6.04

859

9.30

8

Γ3

0.24

6±

0.04

3−

0.05

2±

0.13

30.

0007±

0.00

71−−−

0.70

1±

0.00

4−

0.61

8±

0.15

6−

0.74

6±

0.10

413.6

38±

0.64

058

7.39

860

5.07

8

Γ3a

0.23

5±

0.01

4F

ixed

0.00

19±

0.00

48−−−

0.69

9±

0.00

3−

0.57

9±

0.02

2−

0.71

9±

0.01

413.6

90±

0.21

358

6.04

059

9.30

0

Γ4

0.24

6±

0.04

2−

0.05

2±

0.13

20.

0012±

0.01

30−−−

0.70

1±

0.00

4−

0.61

8±

0.15

5−

0.74

5±

0.10

313.6

38±

0.62

658

7.39

860

5.07

8

Γ4a

0.23

5±

0.01

4F

ixed

0.00

40±

0.00

98−−−

0.69

9±

0.00

3−

0.57

9±

0.02

1−

0.71

9±

0.01

413.6

94±

0.20

758

6.03

159

9.29

1

Γ5

0.24

8±

0.16

2−

0.06

0±

0.21

50.

0060±

0.44

01−−−

0.70

1±

0.00

4−

0.62

4±

0.34

4−

0.74

9±

0.23

013.6

18±

2.78

358

7.42

860

5.10

8

Γ5a

0.24

1±

0.02

3F

ixed

−0.

0050±

0.08

88−−−

0.69

9±

0.00

3−

0.57

0±

0.03

5−

0.71

3±

0.02

313.6

73±

0.49

758

6.49

759

9.75

7

Γ6

0.24

6±

0.04

7−

0.05

3±

0.13

70.

0008±

0.00

97−−−

0.70

1±

0.00

4−

0.61

9±

0.16

4−

0.74

6±

0.10

913.6

34±

0.71

258

7.40

060

5.08

0

Γ6a

0.23

5±

0.01

5F

ixed

0.00

25±

0.00

68−−−

0.69

9±

0.00

3−

0.58

0±

0.02

3−

0.72

0±

0.01

513.6

80±

0.24

358

6.09

359

9.35

3

Γ7

0.24

6±

0.04

3−

0.05

2±

0.13

3−

0.00

08±

0.00

95−−−

0.70

1±

0.00

4−

0.61

8±

0.15

7−

0.74

6±

0.10

513.6

38±

0.64

558

7.39

860

5.07

8

Γ7a

0.23

5±

0.01

5F

ixed

−0.

0028±

0.00

71−−−

0.69

9±

0.00

3−

0.57

9±

0.02

2−

0.71

9±

0.01

513.6

90±

0.21

658

6.04

659

9.30

6

Γ8

0.24

7±

0.03

5−

0.05

1±

0.12

30.

0020±

0.01

80−−−

0.70

1±

0.00

4−

0.61

6±

0.14

2−

0.74

4±

0.09

513.6

46±

0.54

258

7.39

660

5.07

6

Γ8a

0.23

7±

0.01

3F

ixed

0.00

65±

0.01

48−−−

0.69

9±

0.00

3−

0.57

7±

0.02

0−

0.71

8±

0.01

313.7

15±

0.19

558

5.95

059

9.21

0

ωC

DM

0.24

9±

0.02

7−

0.05

9±

0.09

3−−−

−−−

0.70

1±

0.00

4−

0.62

1±

0.10

8−

0.74

8±

0.07

213.6

26±

0.40

958

5.43

359

8.69

3

ΛC

DM

0.24

0±

0.01

4−−−

−−−

−−−

0.69

9±

0.00

3−

0.57

2±

0.02

1−

0.71

4±

0.01

413.6

66±

0.18

958

4.50

259

3.34

2

12

Tab

le4:

Res

ult

sof

the

dat

afitt

ing

usi

ng

the

join

tan

alysi

sfr

omU

nio

n2.

1an

dH

(z).

Mod

elΩ

m0

γx

αβ

hq 0

ωeff

Age

AIC

BIC

Γ1a

0.23

8±

0.08

1F

ixed

0.18

53±

0.45

31−

0.12

07±

0.43

540.

700±

0.00

5−

0.57

4±

0.12

1−

0.71

6±

0.08

113.0

50±

5.70

058

5.80

060

3.44

0

Γ1c

0.15

9±

0.19

60.

100±

0.31

50.

2059±

0.57

26F

ixed

0.70

0±

0.00

5−

0.57

4±

0.46

0−

0.71

6±

0.30

713.0

50±

5.21

458

5.80

060

3.44

0

Γ1d

0.01

0±

0.47

20.

236±

0.41

80.

1954±

0.38

920.

1954±

0.38

920.

700±

0.00

4−

0.58

2±

0.80

2−

0.72

2±

0.53

413.1

63±

7.47

558

5.80

760

3.44

8

Γ1e

0.21

2±

0.05

5F

ixed

Fix

ed0.

0467±

0.12

940.

701±

0.00

4−

0.61

3±

0.08

2−

0.74

2±

0.05

513.6

50±

1.14

958

3.98

559

7.21

5

Γ1f

0.21

7±

0.03

2F

ixed

0.06

50±

0.12

94F

ixed

0.70

1±

0.00

4−

0.60

7±

0.04

8−

0.73

8±

0.03

213.4

15±

1.12

958

3.88

859

7.11

9

Γ1g

0.21

3±

0.04

4F

ixed

0.02

85±

0.06

560.

0285±

0.06

560.

701±

0.00

4−

0.61

2±

0.06

6−

0.74

1±

0.04

413.5

42±

0.91

158

3.93

759

7.16

8

Γ2c

0.15

9±

0.19

60.

100±

0.31

5−

0.17

09±

0.39

38F

ixed

0.70

0±

0.00

5−

0.57

4±

0.46

0−

0.71

6±

0.30

713.0

49±

5.21

658

5.80

060

3.44

0

Γ2d

0.14

0±

0.32

10.

120±

0.43

6−

0.21

02±

0.61

93−

0.21

02±

0.61

930.

700±

0.00

5−

0.57

4±

0.68

0−

0.71

6±

0.45

413.0

51±

7.57

858

5.80

060

3.44

0

Γ2f

0.21

7±

0.03

2F

ixed

−0.

0610±

0.11

41F

ixed

0.70

1±

0.00

4−

0.60

7±

0.04

8−

0.73

8±

0.03

213.4

15±

1.12

858

3.88

859

7.11

9

Γ3

0.17

5±

0.19

40.

079±

0.32

30.

0888±

0.20

04−−−

0.70

0±

0.00

5−

0.57

8±

0.46

4−

0.71

9±

0.30

913.0

25±

4.76

558

5.81

760

3.45

7

Γ3a

0.21

8±

0.02

9F

ixed

0.04

26±

0.08

23−−−

0.70

1±

0.00

4−

0.60

5±

0.04

4−

0.73

7±

0.02

913.3

55±

1.18

158

3.88

059

7.11

0

Γ4

0.20

3±

0.12

50.

039±

0.25

40.

2036±

0.66

99−−−

0.70

0±

0.00

5−

0.58

4±

0.33

8−

0.72

3±

0.22

613.0

12±

4.15

158

5.84

160

3.48

1

Γ4a

0.22

1±

0.02

5F

ixed

0.12

08±

0.22

44−−−

0.70

1±

0.00

4−

0.60

1±

0.03

7−

0.73

4±

0.02

513.2

47±

1.28

758

3.86

859

7.09

8

Γ5a

0.21

3±

0.06

5F

ixed

0.06

84±

0.22

97−−−

0.70

1±

0.00

4−

0.61

3±

0.09

8−

0.74

2±

0.06

513.6

84±

1.26

358

4.02

559

7.25

6

Γ6

0.01

0±

0.29

20.

238±

0.82

00.

1805±

0.69

88−−−

0.70

0±

0.00

5−

0.57

9±

1.20

9−

0.71

9±

0.80

613.1

37±

10.8

2358

5.80

060

3.44

1

Γ6a

0.21

4±

0.04

3F

ixed

0.02

63±

0.05

99−−−

0.70

1±

0.00

4−

0.61

2±

0.06

5−

0.74

1±

0.04

313.5

34±

1.13

358

3.93

359

7.16

4

Γ7

0.13

6±

0.32

60.

126±

0.43

9−

0.18

28±

0.52

64−−−

0.70

0±

0.00

5−

0.57

3±

0.68

8−

0.71

5±

0.45

913.0

59±

6.98

958

5.79

660

3.43

7

Γ7a

0.21

7±

0.03

2F

ixed

−0.

0544±

0.10

89−−−

0.70

1±

0.00

4−

0.60

7±

0.04

9−

0.73

8±

0.03

213.4

20±

1.10

658

3.88

959

7.11

9

Γ8

0.35

5±

0.22

8−

0.19

3±

0.36

20.

2676±

0.56

53−−−

0.70

0±

0.00

5−

0.57

3±

0.52

3−

0.71

5±

0.34

913.0

59±

5.43

858

5.79

660

3.43

7

Γ8a

0.21

8±

0.07

6F

ixed

−0.

0508±

0.27

31−−−

0.70

1±

0.00

5−

0.60

6±

0.11

4−

0.73

7±

0.07

613.8

00±

1.16

658

4.07

759

7.30

8

ωC

DM

0.24

6±

0.04

1−

0.04

7±

0.12

9−−−

−−−

0.70

1±

0.00

4−

0.61

3±

0.15

2−

0.74

2±

0.10

113.6

50±

0.60

758

3.98

559

7.21

5

ΛC

DM

0.23

1±

0.01

6−−−

−−−

−−−

0.70

0±

0.00

3−

0.58

5±

0.02

3−

0.72

3±

0.01

613.7

58±

0.22

158

2.11

859

0.93

8

13

Table 5: Ranking of models according to BIC. In the left panel we show the joint analysisof Union 2.1+H(z)+BAO+CMB and in the right panel we have the joint analysis of binnedJLA+H(z)+BAO+CMB as comparison. f.p. is the number of free parameters in the model.

Model U2.1+BAO+H(z)+CMB bJLA+BAO+H(z)+CMB f.p.ΛCDM 593.348 58.972 2

ωCDM 598.700 62.457 3

Γ8a 599.362 62.346 3

Γ7a 599.367 62.389 3

Γ6a 599.384 62.429 3

Γ1g 599.726 62.543 3

Γ1f 599.729 62.540 3

Γ2f 599.732 62.860 3

Γ4a 599.741 62.544 3

Γ1e 599.767 63.087 3

Γ5a 599.885 63.340 3

Γ3a 599.890 62.542 3

Γ7 605.107 66.359 4

Γ1b 605.108 66.619 4

Γ2b 605.120 66.647 4

Γ1d 605.122 66.358 4

Γ1c 605.123 66.357 4

Γ6 605.126 66.394 4

Γ5 605.135 66.953 4

Γ2c 605.143 66.358 4

Γ2d 605.175 66.482 4

Γ4 605.379 66.358 4

Γ1a 606.137 66.728 4

Γ3 606.009 66.433 4

Γ8 606.991 67.219 4

14

In Table 2 we have also included, besides interacting models, ΛCDM and ωCDM models ascomparison. In this table all interacting scenarios and ωCDM model present a negative value ofthe barotropic index of DE (γx), indicating that there is a trend in favor of phantom DE models.Nevertheless, γx is compatible with zero considering the 1σ confidence level. Besides, we note thatsome of the interacting parameters become smaller than 5 × 10−5 when we include CMB datain the analysis, this is the case for Γ1c and Γ1d. Also, we note that interaction Γ2a is not wellconstrained by the considered data and some of the interactions have a defined sign inside the 1σregion, this is the case of Γ1b, Γ1e, Γ1f , Γ2f , Γ3a, Γ4a, Γ5, Γ5a, Γ6a, Γ7a and Γ8a.

In Table 3 we show the joint analysis considering only Union 2.1+H(z)+BAO, we note thatthe case Γ2a is absent because the error in the β parameter becomes too large (which we canalso observe in Table 2). Here, γx is negative in all the cases and most of the interacting modelshave the same sign in the interacting parameters as in Table 2, but Γ1a, Γ2d, Γ5, Γ5a. Also, incomparing Table 3 to Table 2 we note that interactions Γ1b, Γ1e, Γ5, Γ6a, Γ7a, Γ8 and Γ8a havethe same order of magnitude for interacting parameters when we include CMB data. InteractionsΓ5a, Γ6 and Γ7 increase the values of the interacting parameter and the remaining cases reducetheir absolute value in one or two orders of magnitude when we consider CMB data.

In Table 4 we show the joint analysis considering only Union 2.1 and H(z) data. We note thatmost of interactions have γx > 0, indicating that it is BAO and CMB data which constrain thisparameter to be negative. On the other hand, we do not include in this table interactions Γ1b,Γ2a, Γ2b and Γ5 because the error in the interaction parameters in these cases become too large,as we can see in Table 3 for Γ1b, Γ2b and Γ5 and in Table 2 for Γ2a.

In Tables 2 - 4 we notice that, even though there is a deviation from the ΛCDM scenario,we obtain similar values for the current deceleration parameter q0, the current effective stateparameter ωeff and the age of our universe for all the studied interacting scenarios.

In Table 5 we extend our analysis by considering binned data of the more recent JLA compi-lation of SN Ia [37]. We note that for the joint analysis using Union 2.1 or JLA compilation theresults are consistent, and in light of the Bayesian information criterion, the interacting modelsare ordered according to the number of free parameters of each model.

In our analysis ΛCDM is the model with the lowest AIC and BIC parameters when we use datafrom the joint analysis of Union2.1+H(z)+BAO+CMB (Table 2), Union2.1+H(z)+BAO (Table3), Union2.1+H(z) (Table 4) or binned JLA+H(z)+BAO+CMB (Table 5). From Figure 1 we seethat, when the underlying model is assumed to be ΛCDM, AIC indicates that all models withthree free parameters are in the region of “strong evidence in favor”. Nevertheless under BIC,interacting models with four free parameters are further than having “strong evidence against”and the models of three free parameters are in the upper limit of having “evidence against”. FromFigures 1 and 2, we notice a tension between AIC and BIC results, while AIC indicates there is“evidence in favor” BIC indicates that there is “evidence against” or “strong evidence against”for the same model. This is due to the fact that BIC strongly penalizes models when they have a

15

larger number of parameters [22].Compared to ΛCDM, the studied interacting models have “evidence against”. This is consis-

tent with the results of Ref.[26], where the authors conclude that the particular interacting modelthey study is disfavored compared to ΛCDM, also they notice that BIC is a more restrictivecriteria. The model ωCDM is also incompatible with ΛCDM with respect to BIC.

If we compare the models without considering ΛCDM, the best model according to AIC andBIC is ωCDM when we consider the joint analysis of Union2.1+H(z)+ BAO+CMB. In Table 5we consider only the more stringent criteria, BIC. Here we note that under BIC all models withthree free parameters (f.p.) cannot be ruled out when we assume that ωCDM is the underlyingmodel. In Figure 2 we see that by using BIC there is “strong evidence against” models with 4f.p. when the base model is ωCDM, i.e., we can rule out models of 4 f.p. but not models of 3f.p. if the best model is ωCDM. On the other hand, the best interacting model under BIC (andAIC) is Γ8a, which has an interaction proportional to the deceleration parameter q. Among all ourmodels, those shown in Figure 3 alleviate the coincidence problem, besides, all of them have anenergy transfer from DE to DM today. In the case of Γ8a, for z & 0.7 we have an energy transferfrom DM to DE and for z . 0.7 the energy transfer is from DE to DM as we see in Figure 4.

It is noteworthy to mention that interaction Γ8a is marginally better than other interactingmodels according to AIC and BIC and this interaction alleviates the coincidence problem andchanges sign during evolution. A similar behavior was reported in Ref.[11] where the authorsseparate the data in redshift bins for Q = 3Hδ, where δ is a constant fitted for each bin. Theauthors consider different parametrizations of the equation of state for DE and they found anoscillation of the interaction sign. Sign-changeable interactions have also been studied in Refs.[10]-[12], [27].

As summary, from our analysis we notice that there are consistent interacting models thatexplain the data equally well than ωCDM, and an increase of the number of free parameters ininteracting models, although phenomenologically interesting, is strongly penalized according toBIC in the description of the late universe.

5 Final Remarks

In this work we analyzed eight general types of interacting models of the dark sector with analyticalsolutions and compared how well they fit the joint data from Union 2.1+H(z)+BAO+CMB usingthe Akaike information criterion and the Bayesian information criterion. The main goal of ourwork was to investigate if more complex interacting models (more complex meaning models withmore free parameters) are competitive in fitting the data and whether we could distinguish themvia AIC and BIC.

The models in Table 1 are interesting because they are good candidates to alleviate the co-

16

1a 1b 1c 1d 1e 1f 1g 2a 2b 2c 2d 2f 3 3a 4 4a 5 5a 6 6a 7 7a 8 8a CDM0

2

4

6

8

10

12

14

16Union 2.1 + H(z) + BAO + CMB

BICAIC

Figure 1: ∆AIC and ∆BIC of models defined in Table 2 compared to ΛCDM.

1a 1b 1c 1d 1e 1f 1g 2a 2b 2c 2d 2f 3 3a 4 4a 5 5a 6 6a 7 7a 8 8a0

2

4

6

8

10Union 2.1 + H(z) + BAO + CMB

BICAIC

Figure 2: ∆AIC and ∆BIC of interacting models defined in Table 2 compared to the ωCDMmodel.

incidence problem, furthermore, the physical motivation to the studied models was discussed insection 2, where we showed that the family of interactions presented can be interpreted in terms ofa variable Chaplygin gas in a unified dark sector scenario or in terms of a variable state parameterfor the dark energy component.

Taking into account the theoretical problems that the ΛCDM scenario presents and the ob-servational tensions recently reported with this model [4], we assume that a departure from thesimplest model is needed. We compared a family of interacting models among themselves andwith the ωCDM scenario. In our analysis we noted a tension between the results using AIC andBIC and we decided to follow the more stringent criterion, namely the BIC (Table 5). Accordingto our results, under the BIC “there is not enough evidence against” any interacting model withthree free parameters when we assume that the underlying model is the one which has the lowestBIC parameter, which turns out to be ωCDM. Among the interacting models, Γ8a is the modelwith the lowest BIC parameter value, it corresponds to a sign-changeable interaction with γx = 0and γm = 1 and it is compatible with ωCDM. Furthermore, Γ8a is one of the models that alleviate

17

1.0 0.9 0.8 0.7 0.6 0.5

z10 5

10 4

10 3

10 2

10 1

r(z)

r5(z)r8(z)r8a(z)

CDM

Figure 3: Coincidence parameter in semilog scale. These interactions have an energy transferfrom DE to DM.

0.00 0.25 0.50 0.75 1.00 1.25 1.50 1.75 2.00

z10 5

10 4

10 3

10 2

10 1

100

x

m

| 8a|/

Figure 4: Semilog graphic of the evolution of the density parameters for the interacting modelΓ8a, note that the interaction has a sign change at redshift z ≈ 0.7 approximately.

18

the coincidence problem, since the value of the coincidence parameter in the future tends to aconstant (see Fig. 3).

For the selected models we concluded that all the considered models with three free parametersare compatible among them, i.e. all they have a BIC parameter in the same range, thus thesemodels are not distinguishable, generating in this sense a new kind of degeneracy problem. Asimilar behavior appears when we inspect models with four free parameters as we see in Table 5.Furthermore, it is worth to emphasize that all the interacting models with three free parameters,besides of representing different phenomenology, adjust the data as well as the ωCDM model.

When we compare models with three free parameters to models with four free parameters(using BIC) we find “evidence against” the four free parameters models when we assume that theunderlying model is a three free parameters interacting model.

Finally we conclude that an increase of the complexity of interacting models, measured throughthe number of free parameters, is strongly penalized according to BIC in the description of thelate universe. In the near future we expect to improve this analysis by considering differentparametrizations for the DE state parameter, the dark degeneracy and more sophisticated methodsto constrain data, such as Monte Carlo.

Acknowledgements

This work was partially supported by Direccion de Investigacion de la Universidad del Bıo-Bıothrough grants GI 150407/VC and 151307 3/R. F.A. has been supported by Comision Nacionalde Ciencias y Tecnologıa through Fondecyt Grant 3130736 and the Direccion de Investigacion dela Universidad de La Frontera, Project DI17-0075. We would like to thank Saulo Carneiro for theenlightening comments on the manuscript and Guillermo Rubilar for helpful discussions and thereviewing of the manuscript.

References

[1] A. G. Riess et al. [Supernova Search Team Collaboration], Astron. J. 116, 1009 (1998);S. Perlmutter et al. [Supernova Cosmology Project Collaboration], Astrophys. J. 517, 565(1999).

[2] P. A. R. Ade et al. [Planck Collaboration], Astron. Astrophys. 594, A13 (2016).

[3] E. J. Copeland, M. Sami, S. Tsujikawa, Int. J. Mod. Phys. D 15, 1753 (2006).

[4] A. G. Riess et al., Astrophys. J. 826, no. 1, 56 (2016); A.G. Riess et al., Astrophys. J. 730,119 (2011). Erratum: [Astrophys. J. 732, 129 (2011)].

19

[5] W. Zimdahl, D. Pavon, L. Chimento, Phys. Lett. B 521, 133 (2001).

[6] A. Nunes, J. P. Mimoso, T. C. Charters, Phys. Rev. D 63, 083506 (2001).

[7] G. Mangano, G. Miele, V. Pettorino, Mod. Phys. Lett. A 18, 831 (2003).

[8] W. Zimdahl, D. Pavon, L. P. Chimento, A. S. Jakubi, astro-ph/0404122.

[9] R. Curbelo, T. Gonzalez, G. Leon, I. Quiros, Class. Quantum Gravity 23, 1585 (2006).

[10] H. Wei. Commun. Theor. Phys. 56, 972 (2011).

[11] R. G. Cai, Q. Su, Phys. Rev. D 81, 103514 (2010).

[12] Y. Wang, G. B. Zhao, D. Wands, L. Pogosian, R. G. Crittenden, Phys. Rev. D 92, 103005(2015).

[13] B. Wang, E. Abdalla, F. Atrio-Barandela, D. Pavon, Rep. Prog. Phys. 79, no. 9, 096901(2016).

[14] V. Faraoni, J. B. Dent, E. N. Saridakis, Phys. Rev. D 90, no. 6, 063510 (2014); D. Benistyand E. I. Guendelman, Eur. Phys. J. C 77, no. 6, 396 (2017).

[15] L. P. Chimento, Phys. Rev. D 81, 043525 (2010); L. P. Chimento, AIP Conf. Proc., 147130-38 (2012).

[16] M. Kunz, Phys. Rev. D 80, 123001 (2009).

[17] S. Carneiro, H. A. Borges, JCAP 1406, 010 (2014); H. Velten, H. A. Borges, T. R. P. Carames,Phys. Rev. D 93, no. 6, 063503 (2016).

[18] R. Trotta, Contemp. Phys. 49, 71 (2008).

[19] A. R. Liddle, Annu. Rev. Nucl. Part. Sci. 59, 95 (2009).

[20] H. Akaike, IEEE Trans. Autom. Control, 19 (6) 716-723 (1974).

[21] G. Schwarz, Ann. Stat. 6 (2), 461-464 (1978).

[22] A. R. Liddle, Mon. Not. R. Astron. Soc. 351, L49 (2004).

[23] A. R. Liddle, Mon. Not. R. Astron. Soc. 377, L74 (2007).

[24] W. Godlowski, M. Szydlowski, Phys. Lett. B 623, 10 (2005).

20

[25] D. R. Castro, H. Velten, W. Zimdahl, JCAP 1206, 024 (2012).

[26] M. Szydlowski, A. Krawiec, A. Kurek, M. Kamionka, Eur. Phys. J. C 75, no. 99, 5 (2015).

[27] F. Arevalo, P. Cifuentes, S. Lepe, F. Pena, Astrophys. Space Sci. 352, 899 (2014).

[28] F. Arevalo, A. P. R. Bacalhau, W. Zimdahl, Class. Quantum Gravity 29, 235001 (2012).

[29] D. Pavon, B. Wang, Gen. Relativ. Gravity 41, 1 (2009).

[30] S. del Campo, R. Herrera, D. Pavon, Phys. Rev. D 91, no. 12, 123539 (2015).

[31] Y. Z. Ma, Y. Gong, X. Chen, Eur. Phys. J. C 69, 509 (2010).

[32] V. Salvatelli, N. Said, M. Bruni, A. Melchiorri, D. Wands, Phys. Rev. Lett. 113, no. 18,181301 (2014).

[33] J. H. He, B. Wang, E. Abdalla, Phys. Lett. B 671, 139 (2009); Y. H. Li, X. Zhang, Phys.Rev. D 89, no. 8, 083009 (2014).

[34] Z.K. Guo, Y.Z. Zhang, Phys. Lett. B 645, 326 (2007); Z.K. Guo, Y.Z. Zhang, astro-ph/0509790; G. Sethi, S.K. Singh, P. Kumar, D. Jain, A. Dev, Int. J. Mod. Phys. D 15,1089 (2006); X. Zhang, F. Wu, J. Zhang, JCAP 01 (2006)003; U. Debnath, Astrophys. SpaceSci. 312, 295 (2007); M. Jamil, Int. J. Theor. Phys. 49, 62 (2010).

[35] T. Yang, Z. K. Guo, R. G. Cai, Phys. Rev. D 91, no. 12, 123533 (2015).

[36] N. Suzuki et al., Astrophys. J. 746, 85 (2012).

[37] M. Betoule et al. [SDSS Collaboration], Astron. Astrophys. 568, A22 (2014).

[38] K. Liao, Z. Li, J. Ming, Z. H. Zhu,Phys. Lett. B, 718, 1166 (2013).

[39] C. Blake et al., Mon. Not. R. Astron. Soc. 418, 1707 (2011).

[40] W. J. Percival et al., MNRAS 401, 2148 (2010).

[41] F. Beutler et al., MNRAS 416, 3017 (2011).

[42] N. Aghanim et al. [Planck Collaboration], Astron. Astrophys. 594, A11 (2016).

[43] O. Elgaroy and T. Multamaki, Astron. Astrophys. 471, 65 (2007).

21

[44] S. Carneiro, M. A. Dantas, C. Pigozzo and J. S. Alcaniz, Phys. Rev. D 77, 083504 (2008);C. Pigozzo, M. A. Dantas, S. Carneiro and J. S. Alcaniz, JCAP 1108, 022 (2011).

[45] F. E. M. Costa and J. S. Alcaniz, Phys. Rev. D 81, 043506 (2010); F. E. M. Costa, E. M. Bar-boza, Jr. and J. S. Alcaniz, Phys. Rev. D 79, 127302 (2009); J. F. Jesus, R. C. Santos,J. S. Alcaniz and J. A. S. Lima, Phys. Rev. D 78, 063514 (2008); Q. Wu, Y. Gong, A. Wangand J. S. Alcaniz, Phys. Lett. B 659, 34 (2008).

[46] D. J. Eisenstein et al., Astrophys. J. 633, 560 (2005).

[47] S. Weinberg, Cosmology, (Oxford, 2008).

[48] D. J. Eisenstein and W. Hu, Astrophys. J. 496, 605 (1998).

[49] W. Hu, M. Fukugita, M. Zaldarriaga and M. Tegmark, Astrophys. J. 549, 669 (2001).

[50] W. Hu and N. Sugiyama, Astrophys. J. 471, 542 (1996).

Appendix A:

The distance modulus is defined as

µ(z) := 5 log

[dL(z)

1pc

]− 5, (A1)

where dL(z) is the luminosity distance at redshift z. For a spatially flat universe, we have

dL(z) = (1 + z)r(z) =(1 + z)c

H0

∫ z

0

dz′

E(z′), (A2)

with H0E(z) = H(z), r(z) is the comoving radius at redshift z and c the speed of light.On the other hand, the H(z) dataset is related to the measure of the age difference, ∆t, between

two passively evolving galaxies that formed at the same time but separated by a small redshiftinterval ∆z. One can infer the value of the derivative, (dz/dt), from the ratio (∆z/∆t) [38] andthrough the relation

H(z) = − 1

1 + z

dz

dt, (A3)

infer the value of H for a given z.For BAO’s dataset we need to define the acoustic parameter introduced by Eisenstein and the

BAO typical scale rs(zd), i.e. the comoving radius of the sound horizon at the drag epoch zd,when photons and baryons decouple.

22

The acoustic parameter A(z) is given by [46],

A(z) =DV(z)

√Ωm0H

20

cz, (A4)

with Ωm0 = Ωm0 + Ωb0 where the distance scale DV is defined as

DV(z) =1

H0

[(1 + z)2D2

A(z)cz

E(z)

] 13

, (A5)

and DA(z) is the angular diameter distance,

DA(z) =DL(z)

(1 + z)2, (A6)

with DL(z) = H0dL.Other important function is the dimensionless distance ratio given by

dz(z) =rs(zd)

DV(z), (A7)

where the sound horizon is defined as

rs(z) =

∫ ∞z

cs(z)dz

H(z), (A8)

and the sound speed in the photon-baryon fluid is

cs =c√

3(1 +R), (A9)

where R := 3ρb/4ργ, ρb = ρb0(1 + z)3 is the energy density of baryons and ργ = ργ0(1 + z)4 isthe energy density of photons of the CMB radiation [47]. We use Ωγ0h

2 = 2.469 × 10−5[47] andΩb0h

2 = 0.0222 [2] where Ωγ0 = ργ03H2

0is the normalized energy density of CMB photons today,

Ωb0 = ρb03H2

0is the normalized baryonic energy density today and h is the dimensionless Hubble

parameter such that H0 = 100h km s−1Mpc−1.For the redshift at the drag epoch zd we use the formula proposed by Eisenstein to fit numerical

recombination results [48]:

zd =1291(Ωm0h

2)0.251

1 + 0.659(Ωm0h2)0.828[1 + b1(Ωb0h2)b2 ], (A10)

23

where

b1 = 0.313(Ωm0h2)−0.419[1 + 0.607(Ωm0h

2)0.674],

b2 = 0.238(Ωm0h2)0.223.

From the CMB we use the position of the first peak of the CMB anisotropy spectrum l1 [49]:

l1 = lA(1− δ1) where δ1 = 0.267( r

0.3

)0.1

, (A11)

with r = ρr/(ρm + ρb) evaluated at the redshift of last scattering zls and the radiation densitygiven by [47]:

ρr(z) = 3H20 Ωγ0

(1 +

7

8

(4

11

)4/3

Neff

)(1 + z)4, (A12)

where we have considered the neutrinos’ contribution with Neff = 3.04 [2].The acoustic scale lA is defined as

lA =πdL(zls)

(1 + zls)rs(zls), (A13)

where the last scattering redshift is approximated by [50]:

zls = 1048(1 + 0.00124(Ωb0h

2)−0.738) (

1 + g1(Ωm0h2)g2),

with:

g1 =0.0783(Ωb0h

2)−0.238

1 + 39.5(Ωb0h2)0.763, g2 =

0.560

1 + 21.1(Ωb0h2)1.81.

24