JMPJournal of Modern Physics2153-1196Scientific Research Publishing10.4236/jmp.2021.1212098JMP-112694ArticlesPhysics&Mathematics A New Physics Would Explain What Looks Like an Irreconcilable Tension between the Values of Hubble Constants and Allows <i>H<sub>0</sub></i> to Be Calculated Theoretically Several Ways ClaudeMercier1Independent Researcher, Baie-Comeau, Canada091020211212165617071, September 202123, October 2021 26, October 2021© Copyright 2014 by authors and Scientific Research Publishing Inc. 2014This work is licensed under the Creative Commons Attribution International License (CC BY). http://creativecommons.org/licenses/by/4.0/

Observing galaxies receding from each other, Hubble found the universe’s expansion in 1929. His law that gives the receding speed as a function of distance implies a factor called Hubble constant H0. We want to validate our theoretical value of H0 ≈ 72.09548580(32) km&sdot;s-1&sdot;MParsec-1 with a new cosmological model found in 2019. This model predicts what may look like two possible values of H0. According to this model, the correct equation of the apparent age of the universe gives ~ 14.14 billion years. In approximation, we get the well-known equation 1/H0 ≈ 13.56 billion years. When we force these ages to fit the 1/H0 formula, it gives two different Hubble constant values of ~69.2 and 72.1 km&sdot;s-1&sdot;sdot;MParsec-1. When we apply a theoretical correction factor of η ≈ 1.042516951 on the first value, both target the second one. We found 42 equations of H0 linking different physics constants. Some are used to measure H0 as a function of the average temperature T of the Cosmological Microwave Background and the universal gravitational constant G: H0 ≈ 72.06(90) km&sdot;s-1&sdot;MParsec-1 from T by Cobra probe & Equation (16) H0 ≈ 71.95(50) km&sdot;s-1&sdot;MParsec-1 from T by Partridge & Equation (16) H0 ≈ 72.086(36) km&sdot;s-1&sdot;MParsec-1 from G & Equation (34) H0 ≈ 72.105(36) km&sdot;s-1&sdot;MParsec-1 from G & Equations (74), (75), or (76). With 508 published values, H0 ≈ 72.0957 ± 0.33 km&sdot;s-1&sdot;MParsec-1 seems to be the “ideal” statistical result. It validates our model and our theoretical H0 value which are useful to find various interactions with the different constants. Our model also explains the ambiguity between the different universe’s age measurements and seems to unlock a tension between two H0 values.

Hubble Constant <i>H<sub>0</sub></i> Hubble Tension Age of the Universe
1. Introduction

In astrophysics, the Hubble constant H0  is a parameter to analyze the universe. Nevertheless, it is also one of the lesser-known values.

In 1916, Einstein found the general relativity laws . His equations expect that the universe is either expanding or in a Big Crunch. He could have been the first to predict the universe’s expansion, but influenced by the popular idea, Einstein forced his model to be static with a cosmological constant Λ. In 1922, Friedmann showed from relativity that the universe expands at a calculable rate . In 1927, Georges Lemaitre published independent research , giving what is now known as Hubble’s law. In 1929, Hubble discovered the universe’s expansion . Equation (1) gives Hubble’s law, with v being the receding speed in km·s−1, D being the distance between the observed object and the observer, and H0 being the Hubble constant. He measured about H0 ≈ 500 km·s−1·MParsec−1. His high value was due to a wrong calibration of the cepheids used to evaluate distances. Hubble’s law was correct, but H0 was remaining to be found with accuracy.

v = D H 0 (1)

Physicists get H0 based on far cosmic objects (Cepheids, supernovae, red giants, etc.) or local measurements (CMB, universal gravitational constant G, etc.). Including error margins of published values (see the software in Annex A), H0 is between 19 to 174 km·s−1·MParsec−1. However, two values are often measured ~69.2 and ~72.1 km·s−1·MParsec−1. An irreconcilable tension between some H0 values shows up . Even with good accuracies, their error margins do not always overlap. It may let us think that only one of these values is right. No one considered it possible that both values may be in some way correct.

In 2019, we wrote an article  explaining what may look like two values for H0. We calculated the universe age, obtained a result of complex type, and an apparent age of the universe of ~14.4 billion years. The complete equation may be approximated by 1/H0, giving ~13.56 billion years. We notice that there is a difference of ~4.25% between the approximated and the non-approximated values.

Cosmologists use 1/H0 to calculate the universe’s age. Thus, if we could measure the apparent age of the universe with no approximation, we would conclude wrongly that the Hubble constant is 4.25% lower than it should be.

We hypothesize that two values of H0 are somehow obtained from an approximated and non-approximated equation of the apparent age of the universe. The confusion leads to a tension between two values when there should be only one.

We summarize our cosmological model  to get H0 as a function of α, c, and re. We found ways to measure H0 locally by using the Cosmological Microwave Background (CMB) temperature T and by using the universal gravitational constant G  . Based on our model, we found a theoretical equation to calculate H0 from CODATA values (Committee of Data for Science and Technology) .

H 0 = c α 19 β r e ≈ 72.09548580 ( 32 ) × 10 53     km ⋅ s − 1 ⋅ MParsec − 1 where β = 3 − 5 ≈ 0.76 (2)

We want to validate this theoretical value of H0 and highlight the tension between two measured values of H0. We list the results of the most recent measures of H0 and build a graph showing someway the popularity of each H0 value range.

We list 42 H0 equations. Certain overcome the difficulties to do experimental measurements. We use one of them as a third measurement of H0. Our cosmological model shows that H0 and the speed of light are not constant.

2. Physics Parameters

A compact form of notation is used to display tolerances (i.e., 2.734(10) K means 2.734 ± 0.010 K). The CODATA 2014  is used to compare the results of our new equations with the articles published in 2019 and 2020.

Light speed in a vacuum c = 2 99792458   m ⋅ s − 1

Permeability of free space μ 0 ≈ 4 π × 10 − 7     NA − 2

Permittivity of free space ε 0 ≈ 8.854187817 × 10 − 12     F ⋅ m − 1

Universal gravitational constant G ≈ 6.67408 ( 31 ) × 10 − 11     m 3 ⋅ kg − 1 ⋅ s − 2

Electron rest mass m e ≈ 9.10938356 ( 11 ) × 10 − 31   kg

Classical electron radius r e ≈ 2.8179403227 ( 19 ) × 10 − 15   m

Electron charge q e ≈ − 1.6021766208 ( 98 ) × 10 − 19     C

Planck length L p ≈ 1.616229 ( 38 ) × 10 − 35   m

Planck time t p ≈ 5.39116 ( 13 ) × 10 − 44   s

Planck mass m p ≈ 2 .176470 ( 51 ) × 10 − 8     kg

Planck constant h ≈ 6.626070040 ( 81 ) × 10 − 34     J ⋅ s

Fine-structure constant α ≈ 7.2973525664 ( 17 ) × 10 − 3

Boltzmann constant k b ≈ 1.38064852 ( 79 ) × 10 − 23   J ⋅ K − 1

Rydberg constant R ∞ ≈ 10973731.568508 ( 65 )   m − 1

3. Summary of our Theory

Our theory is based on a cosmological model officially shown in 2019 , but it summarizes papers we wrote at www.pragtec.com/physique since 2011. First, we outline some main milestones as we did in 2020 .

3.1. Our Cosmological Model

We hypothesize that there was one expanding sphere containing all matter at the Big Bang. There was no light. After ~360000 years , electrons became free to move because of a lower density universe, and the light appeared and began to travel through space, creating a 4-D expanding sphere called the “luminous universe”. As the matter cannot travel as fast as light , it created a smaller 4-D expanding sphere, the “material universe”, imbricated in the “luminous universe”.

Einstein found that the presence of a massive object reduces the speed of light vL . Schwarzschild calculated vL in a context of a weak gravitational field Φ using general relativity . With |Φ| << c2 around a spherical mass, Equation (3) gives vL as a function of c and a local refractive index n0 (function of G  ).

v L ( r ) = c n 0 where n 0 = 1 − 2 Φ / c 2 1 + 2 Φ / c 2 and Φ = − G m r ≤ 0 (3)

From an observer on Earth, c seems constant. However, the knowledge of a precise value of c dates only from 19 century . In 1929, Edwin Hubble found that the universe is expanding . As the apparent universe radius increases, the density of this latest must decrease over time, causing the refractive index of the vacuum to drop. As a result, it causes light to accelerate slowly.

In Equation (3), c is the local speed limit for light in a vacuum in our universe area. Admitting that light accelerates while the universe expands, it will tend towards another asymptotical speed limit k affected by a local refractive index n. For now, k is unknown. Let us build Equation (4), which is analog to Equation (3) for the universe . Our universe parcel is at a distance ru from the universe’s apparent mass center mu. The local speed of light c results from Equation (4).

c = k n where n = 1 − 2 Θ / k 2 1 + 2 Θ / k 2 and Θ = − G m u r u ≤ 0 (4)

Similarly to ru, the Ru value is the apparent radius of curvature of the luminous universe   (also called Hubble radius  ). It is a function of c and H0. It is “apparent” since Equation (5) assumes c constant for a time equal to the universe’s age. Now, its speed is c, but it is not constant in our model . It was lower in the past and will increase while the universe expands. The H0 value represents the expansion rate of the material universe in km·s−1·MParsec−1 . It is the local derivative of the velocity of matter v m with respect to the element of distance dr.

H 0 = d v m d r | r = r u = β c r u = β c β R u ⇒ R u = c H 0 (5)

Locally, at a distance r = ru, matter recedes radially from the center of mass of the universe at a rate β times slower than the speed of light c.

r u = β R u = β c H 0 (6)

The apparent mass mu of the universe is given by Equation (7)  :

m u = c 3 G H 0 = R u c 2 G (7)

Our universe parcel is at a distance ru from the center of the mass mu. It travels at a speed vm relative to this latest. The ratio β is the asymptotical speed of light k in a vacuum (when Ru tends towards infinity) influenced by a refractive index n that is itself influenced by a gravitational potential Θ.

v m = β k n where n = 1 − 2 Θ / k 2 1 + 2 Θ / k 2 and Θ = − G m u r u (8)

Hubble measured H0 from the global movement of galaxies at our location , at ru. They have their own movement. As the universe expands, they are generally moving away from each other. The derivative of the material universe speed vm according to the element of distance dr evaluated at r = ru is H0 .

H 0 = d v m d r | r = r u = k β y r u ( 1 ( 1 + y ) 1 − y 2 ) where y = 2 G m u k 2 r u (9)

Solving Equations (4) to (7), and (9) gives Equations (10) to (14) . The expending speed ratio β between the material and the luminous universes is geometric. It is also the ratio between ru and Ru. It is unique to our model and essential to depict many constants and make links between the infinitely large and small in the Dirac hypothesis on large numbers  .

k = c 2 + 5 ≈ 6.17 × 10 8     m ⋅ s − 1 (10)

β = 3 − 5 ≈ 0.764 (11)

R u ≈ 1.28 × 10 26     m (12)

r u ≈ 9.80 × 10 25     m (13)

m u ≈ 1.73 × 10 53     kg (14)

3.2. Our First Method to Measure H<sub>0</sub> as a Function of T (from CMB)

The accuracies of mu, ru, and Ru widely depend on H0 which could be between 19 and 174 km·s−1·MParsec−1 (listed in the software in Annex A). Therefore, a better method of measuring H0 is required to know mu, ru, and Ru more accurately.

We calculated the CMB temperature T as a function of H0 and G . This equation considers the universe as an ideal black body since it would absorb any incident radiation coming from outside, and it does not reflect or transmit any form of energy outside of the luminous universe (since it expands at the speed of light).

T = β k b ( 15 α 2 h 3 c 5 H 0 2 8 π 6 G ) 1 / 4 (15)

Let us isolate H0 from Equation (15). The accuracy mainly depends on the CMB temperature T. Using T ≈ 2.736(17) K (from Cobra probe  ), we get.

H 0 ( T ) = π 3 T 2 k b 2 β 2 α 8 G 15 c 5 h 3 ≈ 72.06 ( 90 )   km ⋅ s − 1 ⋅ MParsec − 1 (16)

with Partridge T ≈ 2.734(10) K, and H0 ≈ 71.95(50) km·s−1·MParsec−1 . As the least accurate value is T, Equation (16) measures H0 from the CMB temperature. These values lead to new links and are in our software in Annex A.

3.3. Dirac Hypothesis about Large Numbers

Dirac found (inaccurately) that large numbers come into a few orders of magnitude with same dimensions quantities ratios  . All ratios come from N, via certain factors . It represents the maximum number of photons in the universe. We get the highest number when the associated mass mph of a photon is the smallest. This happens when the energy of the photons is at its lowest and with a wavelength of the same length as the circumference of the luminous universe (i.e., 2πRu) . Let us calculate mph by equating its corpuscular and wave energies.

m p h c 2 = h c 2 π R u ⇒ m p h = h 2 π R u c ≈ 2.74 × 10 − 69     kg (17)

We get N by dividing the apparent mass mu of the universe (Equation (7)) by the mass mph associated with a photon of 2πRu wavelength (Equation (17)).

N = m u m p h = 2 π c 5 h G H 0 2 ≈ 6.3018 ( 62 ) × 10 121 (18)

If we try to make a precise evaluation of N by using the Equations (6), (7), (16), and (17), we obtain Equation (19) which is dependent mainly on T. We evaluate the result by using the CODATA 2014  and the average CMB temperature from Cobra probe . Finally, we note that N is dimensionless as α.

N = 15 h 2 α 2 β 4 c 10 4 π 5 G 2 k b 4 T 4 ≈ 6.31 ( 15 ) × 10 121 (19)

Assuming α used as a scale factor applied a few times, we postulate Equation (20). It seems impossible to get this equation from standard physics .

POSTULATE#1: N = 1 / α 57 ≈ 6.303419702 ( 84 ) × 10 121 (20)

In the next formulas, Planck temperature is Tp ≈ 1.42 × 1032 K. This is the highest temperature reached at the Big Bang. It happens when we put the entire mass mu in a point-like pellet of Planck length radius Lp. Planck charge is given by qp ≈ 1.88 × 10−18 C.

“Large” numbers are obtained with N exponent a fraction, such as N1/2, N1/3, N1/4, … N1/57, etc. We get these in different ways by using various parameters of the universe . They are always unitless. Some come from Dirac’s hypothesis on large numbers  . Some links will be used later .

N 2 / 3 = m u α m e β 1 / 2 = R u 2 β r e 2 = m p 4 α 4 m e 4 β 2 = m e 2 β m p h 2 α 2 ≈ 1.58 × 10 81 (21)

N 1 / 2 = m p m p h = R u L p = 1 t p H 0 = 2 π T p k b h H 0 = − 1 q e 4 π m u R u α μ 0 ≈ 7.94 × 10 60 (22)

N 1 / 3 = m u r e α m e R u β = m e β m p h α = R u β r e = α q e 2 4 π ε 0 G β m e 2 ≈ 3.99 × 10 40 (23)

N 1 / 4 = T p T ( 15 β 4 α 2 π 3 ) 1 / 4 = k b T m p h c 2 ( π 3 15 β 4 α 2 ) 1 / 4 ≈ 2.82 × 10 30 (24)

N 1 / 6 = r e L p β = m p α m e β = α 3 4 π R ∞ L p β = 2 π r e k b T p h c β ≈ 1.99 × 10 20 (25)

N 1 / 16 = ( 4 π c R ∞ β H 0 ) 57 / 256 = ( β T p T ) 1 / 4 ( 15 α 2 π 3 ) 1 / 16 ≈ 4.10 × 10 7 (26)

N 1 / 19 = 1 4 π R ∞ r e = ( β m e 2 m p h 2 ) 1 / 12 = 16 π 2 β L p R u R ∞ 2 α ≈ 2.57 × 10 6 (27)

N 1 / 57 = q p 2 q e 2 = ( m p 2 m e 2 β ) 1 / 21 = ( q e 2 4 π β ε 0 G m e 2 ) 1 / 20 = 1 α ≈ 137 (28)

In a non-published document , we show over 150 links that give N with various parameters. The universe is well-linked between the infinitely large and the infinitely small. Almost everything changes while the universe is expanding.

3.4. Precise Calculation of H<sub>0</sub>

Unlike Equation (16), we look for an equation that does not use G and T to get H0 since they do not have good accuracies. Usually, G intervenes in the calculations of gravitational force and energy. Without any details (see   ), let us calculate the electrical energy Ee between two electrons separated by a space equal to the classical electron radius re. The electrical energy Ee is not linked to the distance since we get E e = m e c 2 . We evaluate the gravitational energy for the same conditions, finding E g = G m e 2 / r e . If these experiments are done at the luminous universe periphery, we get an electrical energy E ′ e = E e and a gravitational energy E ′ g = E g / β . The ratio between E ′ e and E ′ g gives Equation (29).

E ′ e E ′ g = m e c 2 ( G m e 2 β r e ) = c 2 r e G m e β ≈ 5.45 × 10 42 (29)

As in Equation (20), we realize that the fine-structure constant α plays a role in determining orders of magnitude. By adjusting the exponent of the fine-structure constant α, we obtain a result identical to Equation (29).

1 α 20 ≈ 5.45 × 10 42 (30)

Equations (29) and (30) seam equal. By isolating G, we get an equation that becomes postulate #2. We cannot deduce this equation from standard physics.

POSTULATE#2: G = c 2 r e α 20 m e β ≈ 6.673229809 ( 86 ) × 10 − 11     m 3 ⋅ kg − 1 ⋅ s − 2 where β = 3 − 5 (31)

We associate the wave energy with the energy of the electron mass me.

m e c 2 = h c α 2 π r e (32)

with Equations (20), (31), and (32), we get Equation (33).

H 0 = c α 19 β r e ≈ 72.09548580 ( 32 )   km ⋅ s − 1 ⋅ MParsec − 1 (33)

This value is like Soltis with 72.1 ± 2.0 km·s−1·MParsec−1, Murtinelli’s with 72 . 1 − 1.8 + 2.1 km·s−1·MParsec−1, and Salvatelli’s with 72 . 1 − 2.3 + 3.2 km·s−1·MParsec−1 (see the list of value in our software in Annex A). Our theoretical value seems to make sense.

3.5. Our Second Method to Measure H<sub>0</sub> as a Function of G

We want to find a second way to measure H0 as a function of G. We must use accurate parameters, such as α and the characteristics of the electron (me and re). We look for an equation dependent on G without any rational exponent that reduces the sensitivity. We can use Equations (31) and (33). From each of them, we isolate re, and we make both equal to get H0. Since G is the least precise value, Equation (34) evaluates H0 as a function of G. We used CODATA 2014 values.

H 0 = c 3 α 39 G m e β ≈ 72.086 ( 36 )   km ⋅ s − 1 ⋅ MParsec − 1 (34)

This result is about 25 times more precise than Equation (16) that uses the average CMB temperature T. We included this result in our software in Annex A.

4. Hubble Constant versus the Age of the Universe

We calculate the universe’s age with our cosmological model to understand what seems to be two potential values of H0. We measure H0 by observing cosmological objects. Universe’s age Δtu is of complex type and results from the integral of the inverse of the expanding speed of the material universe vm with the element of distance dr evaluated between the universe’s center of mass (at r = 0) and the apparent material universe radius of curvature at our location ru.

Δ t u = ∫ r = 0 r = r u 1 v m ( r ) d r = Δ t h u + Δ t 0 h (35)

The Δthu is the time elapsed between the horizon (r = rh) and here (r = ru):

Δ t h u = ∫ r = r h r = r u 1 v m ( r ) d r ( Real type result ) (36)

The Δt0h is the elapsed time between r = 0 and the horizon r = rh:

Δ t 0 h = ∫ r = 0 r = r h 1 v m ( r ) d r ( Imaginary type result ) (37)

At the universe horizon r = rh, the speed of light is zero. We cannot see beyond the horizon. The delay Δthu is the time elapsed between the horizon h and our actual position ru in the universe. The delay Δt0h is the time elapsed between the center of mass of the universe and the horizon rh (given by Equation (38)).

r h = 2 G m u k 2 (38)

Performing the integral calculation of Equation (35), we get Equation (39).

∫ 1 v m ( r ) d r = ( z ( r ) + 2 G ⋅ m u ln ( 2 [ k 2 r + z ( r ) ] ) ) β k 3 where     z ( r ) = k 4 r 2 − 4 G 2 m u 2 (39)

We can decompose the age of the universe Δtu into two parts, Δthu and Δt0h. The value Δthu represents the time elapsed between r = rh (at the horizon) and our actual position r = ru in the universe. The value Δt0h gives the time elapsed between r = 0 (at the Big Bang) and r = rh (at the horizon).

Δ t u = Δ t h u + Δ t 0 h ≈ ( 9.50 + 10.47 i ) × 10 9     yearswhere     i = − 1 (40)

The imaginary time Δt0h means that it elapses independently of our time. We cannot see an event between r = 0 and r = rh, and an observer located between r = 0 and rh could not see us. The Δthu equation is:

Δ t h u = 1 H 0 ( ( ω + 2 ln [ ω + β ( 2 + 5 ) ] − ln ( 4 ) ) 2 22 + 10 5 ) ≈ 7 10 H 0 where ω = β 2 ( 9 + 4 5 ) − 4 (41)

The precise equation for Δt0h is:

Δ t 0 h = − ( 2 + π ) H 0 22 + 10 5 i ≈ − 77 100 H 0 i (42)

The modulus of the complex age Δtu gives the universe’s apparent age Tu.

T u = | Δ t u | = | Δ h u + Δ t 0 h | = ( Δ t h u ) 2 + ( Δ t 0 h i ) 2 ≈ 14.14 × 10 9     years (43)

T u = 1 H 0 ( ( ω + 2 ln [ ω + β ( 2 + 5 ) ] − ln ( 4 ) ) 2 22 + 10 5 ) 2 + ( [ − ( 2 + π ) 22 + 10 5 ] i ) 2 ︸ η ≈ 1 ( withabout4 .25%oferror ) (44)

As the square root over the accolade is approximatively equal to 1, we get:

T u ≈ 1 H 0 ≈ 13.56 × 10 9     years (45)

The value of the correction factor between Equations (43) and (45) is η.

η = H 0 T u = H 0 ( Δ t h u ) 2 + ( Δ t 0 h i ) 2 ≈ 1.042516951 (46)

This η explains why scientists currently measure two values of H0. Scientists can only size the apparent age of the universe with different techniques. They cannot measure the real part and the imaginary part of the universe’s age.

There is no “local” or “far” value of H0. There is only one H0. Some techniques give H0 directly, and others need a correction factor. There is no need for any correction factor when H0 is calculated from Equation (33), measured with the CMB temperature with Equation (16), or with the universal gravitational constant G with Equation (34). Other techniques may get similar results than Equation (43), and if we impose that value to fit with Equation (45), we get H ′ 0 .

H ′ 0 = 1 | Δ t u | ≈ 1 14.14 × 10 9     years ≈ H 0 η ≈ 69.2 km ⋅ s − 1 ⋅ MParsec − 1 (47)

However, Equation (45) gives the actual H0 value:

H 0 ≈ 1 13.56 × 10 9     years ≈ 72.1 km ⋅ s − 1 ⋅ MParsec − 1 (48)

If scientists could measure the real part of the universe’s age and associate this value with 1/H0, they would obtain the following value.

H 0 ≈ 1 9.50 × 10 9     years ≈ 102.94 km ⋅ s − 1 ⋅ MParsec − 1 (49)

If scientists could measure the imaginary part somehow, the association of this value with 1/H0 (like in Equation (45)) would give the following H0 value.

H 0 ≈ 1 10.47 × 10 9     years ≈ 93.39 km ⋅ s − 1 ⋅ MParsec − 1 (50)

with different types of experiments to measure the apparent age of the universe, scientists usually get either ~H0 ≈ 69.2 or ~72.1 km·s−1·Mparsec−1. We assume that all calibration factors are used. New techniques could require other unknown corrective factors that have nothing to do with the related phenomenon.

The articles rarely give enough details to check if the process used needs η. Scientists must verify if the η factor is required for their approach.

5. Other Experimental Measurements of Hubble Constant H<sub>0</sub>

In 1929, Hubble made the first observational-based measurements with cepheids and got H0 ≈ 500 km·s−1·MParsec−1 . Sadly, even with a correct principle, his value is higher than the typical value due to errors in distance calibrations.

Let us validate our theoretical H0 with an adequate interpretation of 508 measurements found on the Internet. The ends of their tolerance ranges give 1016 values. To find H0 that has the highest probability to be measured, we compile the number of crossings with the tolerance ranges for each value of H0. It generates a curve with two tips (Figure 1). The higher it is, the greater the chances are that this value of H0 may be part of many tolerance ranges among the collected data.

A simple statistical phenomenon may be described with a Gaussian function. For fitting a wavy curve, it is necessary to make the sum of many Gaussians. A simpler model with fewer degrees of liberty must always be privileged.

A curve fit is done by summing different Gaussians (shown in Figure 2). A better gap fitting reduces the risk of finding other results. Thus, we gave a heavier weight (×10) to all data located between 69.2 and 72.1 km·s−1·MParsec−1 (from our theory). We tried with and without this approach, and it gives about the same result. As it improves the gap fitting, we kept this approach.

Each tip in Figure 1 is approximated in Figure 2 with two positive Gaussians. We force these curves to be around two means, even though there are four positive curves. It removes two degrees of liberty. We must add a negative Gaussian to model the gap between the two mean values. We must elaborate on this negative Gaussian. Our theory predicts “two close values” of H0. On the curve, a deep gap shows up. It is impossible to get such a gap by only adding positives Gaussians which give two little bumps without any gap. To get a real gap, we must add a negative Gaussian. Let us see in Figure 3 what would look like a curve fit without any negative Gaussian. Since the tips are close, they mix up to build only one tip.

The Gaussian sum in Figure 3 peaks around H0 ≈ 71.11 km·s−1·MParsec−1. The result is not close to our theoretical H0 ≈ 72.09548580 km·s−1·MParsec−1 (Equation (33)), but it is about what is found if statistics were used through the whole data set, thinking they should see only one tip. Moreover, Jang & Lee showed a similar value of H0 ≈ 71.17 km·s−1·MParsec−1 (listed in our software in Annex A) that supposedly reduces the tension between the values obtained by cepheids (calibrated on SNe Ia) and CMB.

In Figure 1, we find two groups around H0 ≈ 69.7 and 71.8 km·s−1·MParsec−1. It is known that there is currently a tension between two groups . A significant gap appears between the two tips. The only way to create such a gap is to

withdraw values nearby a specific value. It would then create a negative Gaussian, such as in Figure 2. It is delicate to debate why some values may have been withdrawn. It could be intentional or not. In the past, it was difficult to see a difference between these groups. Now, the tolerances are small enough to clearly see two groups. With recent growing tensions between these two clans, some may be inclined to shrink or shift some tolerance ranges when it overlaps with neighbor values.

In Figure 4, we apply η to the curves around H0 ≈ 69.882 km·s−1·MParsec−1. Then, all curves stand around H0 ≈ 72.36 km·s−1·MParsec−1. Then, with the curves of Figure 4, we build the curve in Figure 5. Figure 6 is a zoom of its tip.

We want to know the precise value of H0 for which the derivative of the Gaussian summation is 0. It corresponds to the highest probability of getting the true H0 value. Unfortunately, the derivative of a Gaussian summation is not an easy equation to get in a software. We rather use a numerical technic to get it. In Figure 6, we show a zoom of the quadratic curve fit around the tip value. Using the equation, we take the derivative and find its maximum. The quadratic equation has the following form:

y ( x ) = A x 2 + B x + C At the tip,the slope is : y ′ ( x ) = d x d y = 2 A x + B = 0 → x = − B 2 A where x = H 0 , y = number of crossings with different tolerance ranges (51)

H 0 = − B 2 A ≈ 2423.2459592464 2 × 16.8057572117 ≈ 72.0957088907   km ⋅ s − 1 ⋅ MParsec − 1 (52)

This result is well centered on our theoretical value within 3 parts per million. Our approach considers that both clans are someway right. Indeed, their different approaches and results also highlight a new phenomenon. It gives credit to our theory of the universe’s complex age that predicts a few possible fake H0 values.

We have 508 data. Each has a tolerance range (that may be symmetrical or not) that generates two H0 values. Therefore, there are a total of imax = 1016 data at the end. The following equation depicts the statistical error et:

e t ≈ ∑ i = 1 i = i max e i 2 i max (53)

We mention that 16 H0 values in our software in Annex A come from statistics. We kept them since some are mixed up with new valuable data information. So, we modify Equation (53) to remove them to reduce their impact on the total et error. We use the following equation where n = 2 × 16 = 32 (each data generates two H0 values) is the number of elements to exclude from our sample. The total et error reduces with the square root of the number of elements included in our sample.

e t ≈ ∑ i = 1 i = i max e i 2 i max ⋅ [ i max i max − n ] (54)

If n = 0, we fall back on Equation (53). With imax = 1016, Equation (53) gives et ≈ ±0.32 km·s−1·MParsec−1. With imax = 1016 and n = 32, Equation (54) rounds up to et ± 0.33 km·s−1·MParsec−1. The impact of these n elements has a very little impact.

H 0 ≈ 72.0957 ± 0.33   km ⋅ s − 1 ⋅ MParsec − 1 (55)

In Annex A, we supply the software used to get this result. All the main steps enumerated in this article are clearly shown. The software uses starting values (found via Excel) to fit the original curve with 5 Gaussian curves (#0 to #4 to use the same numbers as the software). Each Gaussian uses three parameters: μ is the mean value, σ represents the variance, and m is a multiplication factor.

f ( H 0 ) ≈ m σ 2 π e − 1 2 ( H 0 − μ σ ) 2 (56)

Here are the values for the 5 Gaussian curves used to fit the original curve:

For Gaussians #1 and #2, we force the software to use the same mean value. We do the same thing for Gaussians #3 and #4. We also note that the mean value of Gaussian #0 is negative. With these values, we stopped iterating when the sum of squares of errors was lower than 22000. We see in Figure 1 that the obtained approximated curve fits well the original curve. In our software (Annex A), the iterations start with values close to what they should be.

The specificity of our approach is to say that the two clans are someway right. However, we must apply a correction factor to one of them. Indirectly, it gives credit to a complex universe age that predicts a few possible fake values of H0.

After reading this article, scientists should continue their work as they were doing, without applying any correction factor to their raw data. The correction factor should only be used on the final Gaussian curve to analyze data.

6. A Reminder of Different Useful Identities

To avoid repeating everything unnecessarily, we recall different identities that will be used later to determine H0. Planck units are commonly defined as follows.

Plankmass : m p = h c 2 π G ≈ 2.18 × 10 − 8     kg (58)

Planktime : t p = h G 2 π c 5 = L p c ≈ 5.91 × 10 − 44     s (59)

Planklength : L p = h G 2 π c 3 = c t p = 1.61 × 10 − 35     m (60)

PlankTemperature : T p = h c 3 2 π G k b 2 = m p c 2 k b ≈ 1.42 × 10 32     K (61)

Planckcharge : q p = 2 c h ε 0 = − q e α ≈ 1.88 × 10 − 18     C (62)

The fine-structure constant α is linked to Rydberg constant R and the electron mass me by the following equation:

R ∞ = c m e α 2 2 h (63)

The speed of light c is given as a function of μ0 and ε0.

c = 1 μ 0 ε 0 (64)

Associating the mass-energy of a Planck particle with its wave energy and then, using Equations (31), (32), (64), and (62), we get Planck charge qp defined several ways and as a function of c, G, and h like the other Plank units.

q p = 2 h c μ 0 = 4 π m p L p μ 0 = β G h 2 π μ 0 r e 2 c 4 α 19 (65)

The electron’s charge is determined from the mass of the electron me, the classical electron radius re, and the vacuum permeability μ0.

q e = 4 π m e r e μ 0 = − 1.60 × 10 − 19     C (66)

Let us calculate the precise value of the average temperature T of the CMB. We first make equal Equations (16) and (33). Then, we replace G by Equation (31), and we get rid of Plank constant h by its value from Equation (32).

T = m e c 2 k b ( 15 β 6 α 17 π 3 ) 1 / 4 ≈ 2.7367958 ( 16 )   K (67)

This CMB temperature is like Kimura with 2.737 K .

7. Different Equations to Calculate H<sub>0</sub>

For an academic purpose and to show the interdependence of H0 with the other “constants”, we will enumerate equations using various universe parameters. Some overcome the inherent difficulties in measuring H0 and show a roundabout way of obtaining an accurate value of it. We also find some others which depend on interesting values, or more precise ones. Using the constants c, kb, T, me, re, h, G, μ0, ε0, mu, Ru, R, qe, qp, tp, lp, Tp, mp, mph, and β , we find many equations.

The H0 parameter is not constant since 1/H0 represents an approximation of the apparent universe’s age, and H0 get smaller over time. Since the universe is old, H0 changes slowly. If the constancy of all the universe’s parameters is maintained as it is currently done in metrology, the universe’s age and H0 will seem constant.

Results of 508 different experiences reduce the error by 5081/2 ≈ 22.5. It may look like a significant number, but it is nothing besides what has been done to measure the electron characteristics accurately. Particle accelerators use millions of electrons at each experiment, and they repeat these many times to find something new. Computers analyze the collisions’ results to make the electron’s characteristics more and more accurate. It is why there is no manner to get better results than that of Equation (33), as it is based on well-known characteristics of the electron. We will see further many other equations that give precise results.

Replacing G by Equation (31) in Equation (16), we get Equation (68).

H 0 = π 3 k b 2 T 2 α 9 8 r e 15 m e c 3 h 3 β 5 (68)

Replacing h in Equation (68) by using Equation (32), we get Equation (69).

H 0 = k b 2 T 2 π 3 α 21 15 r e 2 m e 4 β 5 c 6 (69)

Replacing h in Equation (16) by using Equation (32), we get Equation (70).

H 0 = k b 2 T 2 β 2 π 3 G α 15 r e 3 m e 3 c 8 (70)

Replacing T in Equation (70) by using Equation (67), we get Equation (71).

H 0 = β α 9 G m e r e 3 (71)

Using Equation (32) in Equation (33), we get Equation (72).

H 0 = 2 π m e c 2 α 18 β h (72)

With Equations (63) and (72), we get the most accurate equation.

H 0 = 4 π c R ∞ α 16 β (73)

Using Equation (31) in Equation (73), we get Equation (74).

H 0 = 4 π G m e R ∞ β 3 / 2 c r e α 4 (74)

Equation (74) gives H0 with G2014 (from CODATA 2014). The measurement of H0(G2014) ≈ 72.105(36) km·s−1·MParsec−1 is our fourth way to measure H0.

Using Equation (63) in Equation (74), we get Equation (75). This equation is also a good candidate for measuring H0 as a function of G.

H 0 = 8 π G h R ∞ 2 β 3 / 2 r e c 2 α 4 (75)

Using Equation (32) in Equation (75), we get Equation (76).

H 0 = 16 π 2 G m e R ∞ 2 β 3 / 2 c α 7 (76)

This equation is another good candidate for measuring H0 as a function of G. The measure still gives the same result as Equation (74).

We will enumerate other equations without making all the rather fastidious demonstrations. However, all these may be found from previous equations.

H 0 = 2 π m p h c 2 h (77)

H 0 = 2 c 2 π R ∞ α 55 G m e (78)

H 0 = β α 9 m e c r e 3 (79)

H 0 = 8 β G m e ( π R ∞ ) 3 α 9 (80)

H 0 = 8 c β r e ( π R ∞ ) 3 α 29 (81)

H 0 = 2 π c 5 α 57 h G (82)

H 0 = 2 c π β R ∞ α 25 r e (83)

H 0 = 2 π β m e c 3 α 37 h r e (84)

H 0 = 2 π 2 h ( k b 4 T 4 15 m u α 2 c 2 β 4 ) 1 / 3 (85)

H 0 = 2 π k b T p h α 57 (86)

H 0 = c μ 0 q e 2 α 19 β 4 π m e r e 2 (87)

H 0 = q e 2 α 19 β 4 π c ε 0 m e r e 2 (88)

H 0 = q p 2 α 18 β 2 h ε 0 r e (89)

H 0 = c μ 0 q p 2 α 20 β 4 π m e r e 2 (90)

H 0 = q p 2 α 20 β 4 π c ε 0 m e r e 2 (91)

H 0 = c β q p 2 α 19 4 π ε 0 k b T p r e 2 (92)

H 0 = c β q e 2 α 17 4 π ε 0 k b T p r e 2 (93)

H 0 = c q e 2 β 2 4 ε 0 k b T r e 2 ( 15 α 97 π 7 ) 1 / 4 (94)

H 0 = 2 k b T β h ( π 7 α 55 15 ) 1 / 4 (95)

H 0 = k b T β r e m e c ( π 3 α 59 15 ) 1 / 4 (96)

H 0 = 4 π k b T β c μ 0 q e 2 ( π 3 α 59 15 ) 1 / 4 (97)

H 0 = 4 π k b T β c μ 0 q p 2 ( π 3 α 55 15 ) 1 / 4 (98)

H 0 = G k b T c 3 r e 2 ( π 3 15 α 21 ) 1 / 4 (99)

H 0 = 2 G m e k b T h r e c 2 ( π 7 15 α 25 ) 1 / 4 (100)

H 0 = G m p k b T m e c 3 r e 2 ( π 3 α 21 15 β 2 ) 1 / 4 (101)

H 0 = G m u k b T m e c 3 r e 2 ( π 3 α 135 15 β 2 ) 1 / 4 (102)

H 0 = G m e k b T m p c 3 r e 2 ( π 3 β 2 15 α 63 ) 1 / 4 (103)

H 0 = 16 c G k b T ε 0 m e 2 q e 4 ( π 11 15 α 21 ) 1 / 4 (104)

H 0 = β m e 2 8 π 3 G c 3 α 15 h 3 (105)

H 0 = c 3 α 39 G m e β (106)

The last equation measures H0 from G since all other constants are accurate. Many equations are excellent candidates for measuring H0 as a function of G or T. These equations could represent valuable tools for cosmologists.

This document gives 42 equations of H0 as a function of various universe parameters. Since H0 may be defined using different parameters, we suggest that some of the most critical universe parameters are well linked, as much in the infinitely small as in the infinitely large, and H0 is part of these.

8. Why Is H<sub>0</sub> Not Really a Constant?

We want to explain why Hubble parameter H0 cannot be constant over time. As simple as it is, the reverse of Hubble parameter H0 is related to the apparent age of the universe (see Equation (45)). Consequently, the H0 parameter is changing over time. It is, therefore, by abuse of language that we call H0 the Hubble “constant”. To be more precise, we should say the Hubble “parameter”.

When H0 is expressed in km·s−1·MParsec−1, the ninth digit after the dot changes every year. It goes completely unnoticed. More than that, even if we could achieve this precision in our measurements of H0, it would still go unnoticed since we forced c to be constant in 1983. In metrology, scientists choose the speed of light as a standard. Even though c changes every year, if we force it to be constant, we willfully readjust all other constants and units (distance, time, and mass) as a function of c to keep it constant. Then, H0 looks constant as other parameters.

9. Conclusions

This article aimed to show that our theoretical value from Equation (33) (giving H0 ≈ 72.09548580(32) km·s−1·MParsec−1)  is the right one, despite a growing tension  between values around 69.2 and 72.1 km·s−1·MParsec−1.

With 508 data (from  to  shown in our software in Annex A), a graph showing the actual tension  between two values is shown. We decomposed the curve into Gaussians. A negative one is required to explain the large gap between the two H0 values, and it is due to withdrawn values. So, we restored them by removing that curve. Then, we applied a η ≈ 1.042516951 correction factor (from our theory) to the curves located at ~H0 ≈ 69.2 km·s−1·MParsec−1. Our theory highlights a misunderstanding of the link between 1/H0 and the universe’s apparent age. With the proper correction factor applied, we get a statistical value of H0 ≈ 71.85 ± 0.33 km·s−1·MParsec−1, which is close to our theoretical value. Our discovery of the η factor may help to reduce the tension between scientists. Someway we show that even if two H0 values seem to be commonly found with various techniques, both are accurate if a proper correction factor is used.

With a new cosmological model, we get an apparent age of the universe of about 14.14 billion years. The exact formula is approximated from an elaborate integral result by the well-known 1/H0 equation that gives 13.56 billion years. Different techniques may lead to either value. It depends if it is an attempt to measure the universe’s age locally or far away. There is no “local” or “distant” value of H0, as some may pretend  . Sticking their measurement of the apparent age of the universe to 1/H0, most cosmologists get results that stand around 69.2 or 72.1 km·s−1·MParsec−1. Our hypothesis may explain the actual tension  relative to these two values. However, there is only one true H0 value, and the other one is just misinterpreted as being the Hubble constant without quite being so.

Even if many theoretical equations of H0 are shown in this article, we highlight that we also found a few interesting ways to measure the H0 accurately using the CMB temperature T and the value of the universal gravitational constant G from CODATA 2014. These results confirm our theoretical value.

H0 ≈ 72.06(90) km·s−1·MParsec−1 and 71.95(50) km·s−1·MParsec−1 from T,

H0 ≈ 72.086(36) km·s−1·MParsec−1 and 72.105(36) km·s−1·MParsec−1 from G.

For an academic purpose, we enumerated 42 equations of H0 using different parameters. These equations showed that H0 is intricated with all other “constants”. For metrology purposes, the speed of light in a vacuum is forced to be constant to be an unchanging standard. If this situation is considered valid in a metrology context, H0 should also be considered constant and become part of the CODATA. For the same reasons, H0 must also be constant in a metrology context and become part of the CODATA. However, if 1/H0 represents an approximation of the universe’s age, it would also make sense to say that H0 is changing over time.

Einstein’s and Schwarzschild’s equations show that massive objects such as the universe influence the speed of light. As the universe expands, its density diminishes, and the local speed of light increases over time.

The fine-structure constant α is unitless and may be described as a ratio where the variation rate at the numerator counterbalances the variation rate at the denominator. Apart from α and β , all “constants” used to describe H0 in our equations somehow emanate from fundamental units such as the meter, the second, and the kilogram. These units are now defined by the speed of light. As H0 describes the universe’s age and depends on many unit-dependent “constants” based on c, we should consider c and all universe’s unit-dependent parameters as changing over time. Forcing c to be constant is necessary for metrology purposes, but it is not in the interest of physicists for explaining phenomena. An accurate value of H0 has a great interest in deepening our understanding of the universe.

Conflicts of Interest

The author claims that he has no conflict of interest in connection with the publication of this article.

Cite this paper

Mercier, C. (2021) A New Physics Would Explain What Looks Like an Irreconcilable Tension between the Values of Hubble Constants and Allows H0 to Be Calculated Theoretically Several Ways. Journal of Modern Physics, 12, 1656-1707. https://doi.org/10.4236/jmp.2021.1212098

Annex A (C++ Software)

// This software finds "the best" experimental value of H0 with a set of 508 data

// https://sourceforge.net/projects/orwelldevcpp/

#include

#include

#include

#define printf __mingw_printf

#define nbH0 508 //Number of measurements of H0 analyzed

#define Pi 3.141592654 //Definition of Pi

double Mean,Sigma,Multiplier;//Caracteristics of Gaussian curves

double A,B,C; //Coefficients of the quadratic equation of the tip

double LMSTip; //Least mean square for the approximation tip curve

double BEH; //Best estimate of H0

double H0[2*nbH0-1],NbCrossings[2*nbH0-1];

double nbCrossings[2*nbH0-1]; //Vector of number of crossings

double Accuracy_ppm; //Accuracy of H0 compared to the theoretical value

double TVH = 72.09548580; //Theoretical value of H0 (in km/(s*MParsec)

unsigned int PosTipIndex; //Index corresponding to the tip of nbCrossing array

int n = 20; //Sample before & after PosTipIndex to build the tip equation

double Sqr(double value) {return value*value;} //***Returns the square value

//***Function that returns the square root value

double Sqrt(double Value) {

double D; /*Dummy value*/ double V; /*Returned value*/ int i; /*Counter*/

V=0; D=Value;

for (i=0;i<=50;i++) { //Gives 50 bits of precision

D=-D/2;

if (D<0) while (V*V>Value) V=V+D; else while (V*V

return V;

} //End of Sqrt

double Exp(double Value) { return pow(2.718281828,Value);} //***e^Value

//***This procedure creates a table of 2*nbH increasing values

//***of H0[i] with the tolerances T[i] and the signs +/- Variation[i].

void CreateTableOfCrossingH0Ranges(void) {

double PT[nbH0],NT[nbH0]; //Positive tolerance and negative tolerance

double Variation[2*nbH0]; //Variation from the tolerance range

double DH0,DSgn; //Dummy H0 and Sgn used to put H in increasing order

int DeltaCrossings; //Variation of on the number of crossings

int i,j,k; //Counters

i=0;

//We enumerate all H0 values and their tolerance range found on Internet

//In brackets, we add the bibliographic references

H0[i]=69; PT[i]=+16; NT[i++]=-8; //  Abbott et al.

H0[i]=70; PT[i]=+2.7; NT[i++]=-2.7; //  Addisson

H0[i]=72.4; PT[i]=+3.9; NT[i++]=-4.8; //  Addisson

H0[i]=73.1; PT[i]=+3.3; NT[i++]=-3.9; //  Addisson

H0[i]=73.2; PT[i]=+1.3; NT[i++]=-1.3; //  Addisson

H0[i]=68.7; PT[i]=+1.3; NT[i++]=-1.3; //  Addisson

H0[i]=73.5; PT[i]=+5.3; NT[i++]=-5.3; //  Baxter & Sherwin

H0[i]=73.3; PT[i]=+0.7; NT[i++]=-0.7; //  Blakeslee et al.

H0[i]=73.78; PT[i]=+0.84; NT[i++]=-0.84; //  Bonilla

H0[i]=73.577; PT[i]=+0.106; NT[i++]=-0.106; //  Dainotti et al.

H0[i]=73.493; PT[i]=+0.144; NT[i++]=-0.144; //  Dainotti et al.

H0[i]=73.222; PT[i]=+0.262; NT[i++]=-0.262; //  Dainotti et al.

H0[i]=73.664; PT[i]=+0.223; NT[i++]=-0.223; //  Dainotti et al.

H0[i]=73.576; PT[i]=+0.105; NT[i++]=-0.105; //  Dainotti et al.

H0[i]=73.513; PT[i]=+0.142; NT[i++]=-0.142; //  Dainotti et al.

H0[i]=73.192; PT[i]=+0.265; NT[i++]=-0.265; //  Dainotti et al.

H0[i]=73.678; PT[i]=+0.223; NT[i++]=-0.223; //  Dainotti et al.

H0[i]=71.8; PT[i]=+3.9; NT[i++]=-3.3; //  Denzel et al.

H0[i]=72.94; PT[i]=+0.75; NT[i++]=-0.75; //  Di Valentino

H0[i]=72.7; PT[i]=+1.1; NT[i++]=-1.1; //  Di Valentino

H0[i]=68.8; PT[i]=+45.7; NT[i++]=-25.5; //  Gayathri et al.

H0[i]=62.3; PT[i]=+9.1; NT[i++]=-9.1; //  Hagstotz et al.

H0[i]=70.5; PT[i]=+2.37; NT[i++]=-2.37; //  Kethan et al.

H0[i]=72.86; PT[i]=+0.036; NT[i++]=-0.036; //Mercier (this document)

H0[i]=72.105; PT[i]=+0.036; NT[i++]=-0.036; //Mercier (this document)

H0[i]=68.3; PT[i]=+4.6; NT[i++]=-4.6; //  Mukherjee et al.

H0[i]=70; PT[i]=+0.5; NT[i++]=-0.5; //  Park et al.

H0[i]=65.1; PT[i]=+3; NT[i++]=-5.4; //  Philcox et al.

H0[i]=65.6; PT[i]=+3.4; NT[i++]=-3.5; //  Philcox et al.

H0[i]=70.6; PT[i]=+3.7; NT[i++]=-5; //  Philcox et al.

H0[i]=78.3; PT[i]=+2.9; NT[i++]=-2.9; //  Qi et al.

H0[i]=73.6; PT[i]=+1.8; NT[i++]=-1.6; //  Qi et al.

H0[i]=73; PT[i]=+1.4; NT[i++]=-1.4; //  Riess et al.

H0[i]=73.2; PT[i]=+1.3; NT[i++]=-1.3; //  Riess et al.

H0[i]=72.1; PT[i]=+2; NT[i++]=-2; //  Soltis et al.

H0[i]=69.5; PT[i]=+4; NT[i++]=-4; //  Wang & Giannios

H0[i]=71; PT[i]=+20; NT[i++]=-20; //  Zhang et al.

H0[i]=67.4; PT[i]=+0.5; NT[i++]=-0.5; //  Aghanim et al.

H0[i]=67.73; PT[i]=+0.41; NT[i++]=-0.41; //  Benevento

H0[i]=68.22; PT[i]=+0.39; NT[i++]=-0.39; //  Benevento

H0[i]=72.5; PT[i]=+1.85; NT[i++]=-1.85; //  Benevento

H0[i]=69.17; PT[i]=+1.09; NT[i++]=-1.09; //  Benevento

H0[i]=74.5; PT[i]=+5.6; NT[i++]=-6.1; //  Birrer et al.

H0[i]=67.4; PT[i]=+4.1; NT[i++]=-3.2; //  Birrer et al.

H0[i]=75.35; PT[i]=+1.68; NT[i++]=-1.68; //  Camarena & Marra

H0[i]=74; PT[i]=+0.625; NT[i++]=-0.625; //  Chang & Zhu

H0[i]=73.8; PT[i]=+6.3; NT[i++]=-5.8; //  Coughlin et al.

H0[i]=71.2; PT[i]=+3.2; NT[i++]=-3.1; //  Coughlin et al.

H0[i]=72.4; PT[i]=+1.4; NT[i++]=-1.4; //  D'Agostino

H0[i]=71.5; PT[i]=+1.3; NT[i++]=-1.3; //  D'Agostino

H0[i]=71.54; PT[i]=+1.78; NT[i++]=-1.78; //  Dai WM et al.

H0[i]=73.12; PT[i]=+1.14; NT[i++]=-1.14; //  Dai WM et al.

H0[i]=66.2; PT[i]=+4.4; NT[i++]=-4.2; //  Dietrich et al.

H0[i]=69.9; PT[i]=+0.84; NT[i++]=-0.86; //  Gonzalez et al.

H0[i]=71; PT[i]=+4; NT[i++]=-4; //  González-Serrena et al.

H0[i]=74.62; PT[i]=+12.35; NT[i++]=-11.34; //  Haboury

H0[i]=71.89; PT[i]=+11.02; NT[i++]=-10.17; //  Haboury

H0[i]=76.44; PT[i]=+55.76; NT[i++]=-50.17; //  Haboury

H0[i]=50.9; PT[i]=+31.1; NT[i++]=-31.9; //  Haboury

H0[i]=50.81; PT[i]=+28.19; NT[i++]=-27.81; //  Haboury

H0[i]=71; PT[i]=+2; NT[i++]=-3; //  Harvey

H0[i]=65.9; PT[i]=+1.5; NT[i++]=-1.5; //  Holanda et al.

H0[i]=65.9; PT[i]=+4.4; NT[i++]=-4; //  Holanda et al.

H0[i]=64.3; PT[i]=+4.5; NT[i++]=-4.4; //  Holanda et al.

H0[i]=66.8; PT[i]=+13.4; NT[i++]=-9.2; //  Howlett & Davis

H0[i]=64.8; PT[i]=+7.3; NT[i++]=-7.2; //  Howlett & Davis

H0[i]=75.8; PT[i]=+5.2; NT[i++]=-4.9; //  Jaeger et al.

H0[i]=65.8; PT[i]=+3.5; NT[i++]=-3.5; //  Kim et al.

H0[i]=72.3; PT[i]=+1.4; NT[i++]=-1.4; //  Kreisch et al.

H0[i]=71.5; PT[i]=+11.9; NT[i++]=-10.6; //  Li & Zhang

H0[i]=74.7; PT[i]=+5.8; NT[i++]=-5.8; //  Lombriser

H0[i]=72.06; PT[i]=+0.09; NT[i++]=-0.09; //  Mercier

H0[i]=74; PT[i]=+1.6; NT[i++]=-1.6; //  Millon et al.

H0[i]=74.2; PT[i]=+1.7; NT[i++]=-1.8; //  Millon et al.

H0[i]=50.4; PT[i]=+28.1; NT[i++]=-19.5; //  Mukherjee et al.

H0[i]=62.2; PT[i]=+29.5; NT[i++]=-19.7; //  Mukherjee et al.

H0[i]=43.1; PT[i]=+24.6; NT[i++]=-11.4; //  Mukherjee et al.

H0[i]=67.6; PT[i]=+4.3; NT[i++]=-4.2; //  Mukherjee et al.

H0[i]=68.6; PT[i]=+14; NT[i++]=-8.5; //  Nicolaou et al.

H0[i]=69.6; PT[i]=+1; NT[i++]=-1.3; //  Niedermann and Sloth

H0[i]=71.4; PT[i]=+1; NT[i++]=-1; //  Niedermann and Sloth

H0[i]=72; PT[i]=+12; NT[i++]=-8.2; //  Palmese et al.

H0[i]=69.03; PT[i]=+0.87; NT[i++]=-0.87; //  Pandey et al.

H0[i]=70.6; PT[i]=+1.1; NT[i++]=-1.1; //  Pandey et al.

H0[i]=68.44; PT[i]=+0.52; NT[i++]=-0.52; //  Pandey et al.

H0[i]=68.1; PT[i]=+0.58; NT[i++]=-0.58; //  Pandey et al.

H0[i]=73.9; PT[i]=+3; NT[i++]=-3; //  Pesce et al.

H0[i]=68.6; PT[i]=+1.8; NT[i++]=-1.8; //  Pogosian et al.

H0[i]=74.03; PT[i]=+1.42; NT[i++]=-1.42; //  Rui-Yun et al.

H0[i]=75.1; PT[i]=+2.3; NT[i++]=-2.3; //  Schombert et al.

H0[i]=74.2; PT[i]=+2.7; NT[i++]=-3; //  Shajib et al.

H0[i]=67.52; PT[i]=+0.96; NT[i++]=-0.95; //  Sharov & Sinyakov

H0[i]=70.87; PT[i]=+1.63; NT[i++]=-1.62; //  Sharov & Sinyakov

H0[i]=69; PT[i]=+29; NT[i++]=-14; //  Vasylyev & Filippenko

H0[i]=67; PT[i]=+41; NT[i++]=-26; //  Vasylyev & Filippenko

H0[i]=71; PT[i]=+34; NT[i++]=-30; //  Vasylyev & Filippenko

H0[i]=70; PT[i]=+29; NT[i++]=-18; //  Vasylyev & Filippenko

H0[i]=72.3; PT[i]=+2.9; NT[i++]=-2.8; //  Vogl

H0[i]=75.3; PT[i]=+3; NT[i++]=-2.9; //  Wei & Melia

H0[i]=75.3; PT[i]=+1.9; NT[i++]=-1.9; //  Wei & Melia

H0[i]=67.9; PT[i]=+1.1; NT[i++]=-1.3; //  Wu et al.

H0[i]=72; PT[i]=+2.1; NT[i++]=-2.5; //  Wu et al.

H0[i]=73.65; PT[i]=+1.95; NT[i++]=-2.26; //  Yang et al.

H0[i]=67.95; PT[i]=+0.78; NT[i++]=-1.03; //  Zhang & Huang

H0[i]=69.81; PT[i]=+2.22; NT[i++]=-2.7; //  Zhang & Huang

H0[i]=66.75; PT[i]=+3.42; NT[i++]=-4.23; //  Zhang & Huang

H0[i]=70.75; PT[i]=+1.55; NT[i++]=-1.55; //  Agrawal

H0[i]=73.7; PT[i]=+1.4; NT[i++]=-1.4; //  Anderson

H0[i]=72.5; PT[i]=+2.1; NT[i++]=-2.3; //  Birrer

H0[i]=67.4; PT[i]=+0.5; NT[i++]=-0.5; //  Chang et al.

H0[i]=82.8; PT[i]=+9.4; NT[i++]=-8.3; //  Chen et al.

H0[i]=70.1; PT[i]=+5.3; NT[i++]=-4.5; //  Chen et al.

H0[i]=77; PT[i]=+4; NT[i++]=-4.6; //  Chen et al.

H0[i]=75.6; PT[i]=+3.2; NT[i++]=-3; //  Chen et al.

H0[i]=76.8; PT[i]=+2.6; NT[i++]=-2.6; //  Chen et al.

H0[i]=75.7; PT[i]=+4.5; NT[i++]=-4.4; //  Collett

H0[i]=76.8; PT[i]=+4.2; NT[i++]=-3.8; //  Collett

H0[i]=74.2; PT[i]=+3; NT[i++]=-2.9; //  Collett

H0[i]=67.6; PT[i]=+1.1; NT[i++]=-1.1; //  Cuceu et al.

H0[i]=67.4; PT[i]=+6; NT[i++]=-6.2; //  Domínguez

H0[i]=66.6; PT[i]=+1.6; NT[i++]=-1.6; //  Domínguez

H0[i]=70.3; PT[i]=+1.36; NT[i++]=-1.35; //  Dutta et al.

H0[i]=77; PT[i]=+37; NT[i++]=-18; //  Fishbach et al.

H0[i]=76; PT[i]=+19; NT[i++]=-13; //  Fishbach et al.

H0[i]=69.8; PT[i]=+0.8; NT[i++]=-0.8; //  Freedman et al.

H0[i]=68.09; PT[i]=+0.45; NT[i++]=-0.45; //  Guo et al.

H0[i]=69.34; PT[i]=+0.93; NT[i++]=-0.93; //  Guo et al.

H0[i]=69.67; PT[i]=+0.95; NT[i++]=-0.94; //  Guo et al.

H0[i]=69.36; PT[i]=+0.82; NT[i++]=-0.82; //  Guo et al.

H0[i]=69.25; PT[i]=+0.99; NT[i++]=-0.99; //  Guo et al.

H0[i]=74; PT[i]=+16; NT[i++]=-8; //  Hotokezaka et al.

H0[i]=70.3; PT[i]=+5.3; NT[i++]=-5; //  Hotokesaka et al.

H0[i]=82.4; PT[i]=+8.4; NT[i++]=-8.3; //  Jee et al.

H0[i]=67; PT[i]=+3; NT[i++]=-3; //  Kozmanyan et al.

H0[i]=72.2; PT[i]=+2.1; NT[i++]=-2.1; //  Liao et al.

H0[i]=73; PT[i]=+2.8; NT[i++]=-3; //  Liao et al.

H0[i]=67.8; PT[i]=+1.3; NT[i++]=-1.3; //  MacAulay et al.

H0[i]=67.37; PT[i]=+0.62; NT[i++]=-0.62; //  Martinelli

H0[i]=68.8; PT[i]=+1.6; NT[i++]=-1.6; //  Martinelli

H0[i]=73.9; PT[i]=+2.5; NT[i++]=-2.5; //  Martinelli

H0[i]=67.68; PT[i]=+0.46; NT[i++]=-0.46; //  Martinelli

H0[i]=68.4; PT[i]=+1; NT[i++]=-1; //  Martinelli

H0[i]=69.2; PT[i]=+1.5; NT[i++]=-1.5; //  Martinelli

H0[i]=67.51; PT[i]=+0.61; NT[i++]=-0.61; //  Martinelli

H0[i]=68.9; PT[i]=+1.1; NT[i++]=-1.1; //  Martinelli

H0[i]=72.1; PT[i]=+2.1; NT[i++]=-1.8; //  Martinelli

H0[i]=67.75; PT[i]=+0.46; NT[i++]=-0.46; //  Martinelli

H0[i]=68.59; PT[i]=+0.86; NT[i++]=-0.86; //  Martinelli

H0[i]=69.6; PT[i]=+1.3; NT[i++]=-1.3; //  Martinelli

H0[i]=71.505; PT[i]=+0.03; NT[i++]=-0.03; //  Mercier

H0[i]=69; PT[i]=+1.7; NT[i++]=-1.7; //  Park & Ratra

H0[i]=69.8; PT[i]=+1.8; NT[i++]=-1.8; //  Park & Ratra

H0[i]=68.9; PT[i]=+1.7; NT[i++]=-1.7; //  Park & Ratra

H0[i]=70.1; PT[i]=+1.9; NT[i++]=-1.9; //  Park & Ratra

H0[i]=68.5; PT[i]=+1.8; NT[i++]=-1.8; //  Park & Ratra

H0[i]=69.6; PT[i]=+1.9; NT[i++]=-1.9; //  Park & Ratra

H0[i]=72; PT[i]=+1.9; NT[i++]=-1.9; //  Reid

H0[i]=73.5; PT[i]=+1.4; NT[i++]=-1.4; //  Reid

H0[i]=74.22; PT[i]=+1.82; NT[i++]=-1.82; //  Riess et al.

H0[i]=74.03; PT[i]=+1.42; NT[i++]=-1.42; //  Riess et al.

H0[i]=72.8; PT[i]=+1.1; NT[i++]=-1.1; //  Riess

H0[i]=74.3; PT[i]=+1; NT[i++]=-1; //  Riess

H0[i]=71.6; PT[i]=+3.8; NT[i++]=-4.9; //  Rusu et al.

H0[i]=67.99; PT[i]=+0.91; NT[i++]=-0.88; //  Ryan

H0[i]=68.24; PT[i]=+2.39; NT[i++]=-2.33; //  Ryan

H0[i]=66.79; PT[i]=+2.6; NT[i++]=-2.32; //  Ryan

H0[i]=66.8; PT[i]=+2.5; NT[i++]=-2.3; //  Ryan

H0[i]=66.13; PT[i]=+1.38; NT[i++]=-2.09; //  Ryan

H0[i]=67.1; PT[i]=+2.4; NT[i++]=-2.3; //  Ryan

H0[i]=68.44; PT[i]=+0.7; NT[i++]=-0.69; //  Ryan

H0[i]=69.32; PT[i]=+1.42; NT[i++]=-1.42; //  Ryan

H0[i]=68; PT[i]=+2.27; NT[i++]=-1.94; //  Ryan

H0[i]=66.6; PT[i]=+2.2; NT[i++]=-1.9; //  Ryan

H0[i]=67.19; PT[i]=+1; NT[i++]=-1.6; //  Ryan

H0[i]=66.8; PT[i]=+1.8; NT[i++]=-1.7; //  Ryan

H0[i]=63.13; PT[i]=+6.48; NT[i++]=-6.48; //  Saha & Sahoo

H0[i]=74.2; PT[i]=+2.7; NT[i++]=-3; //  Shajib et al.

H0[i]=75; PT[i]=+40; NT[i++]=-32; //  Soares-Santos

H0[i]=78; PT[i]=+96; NT[i++]=-24; //  Soares-Santos

H0[i]=73.1; PT[i]=+0.7; NT[i++]=-0.7; //  Taubenberger et al.

H0[i]=68; PT[i]=+14; NT[i++]=-7; //  Tiwari et al.

H0[i]=68; PT[i]=+18; NT[i++]=-8; //  Tiwari et al.

H0[i]=73.9; PT[i]=+1; NT[i++]=-1; //  Verde et al.

H0[i]=72.5; PT[i]=+1.2; NT[i++]=-1.2; //  Verde et al.

H0[i]=73.3; PT[i]=+1.7; NT[i++]=-1.8; //  Wong et al.

H0[i]=72.4; PT[i]=+2; NT[i++]=-2; //  Yuan et al.

H0[i]=68.36; PT[i]=+0.53; NT[i++]=-0.52; //  Zhang & Huang

H0[i]=64.9; PT[i]=+4.6; NT[i++]=-4.3; //  Zeng and Yan

H0[i]=67.4; PT[i]=+1.1; NT[i++]=-1.2; //  Abbott et al.

H0[i]=69.3; PT[i]=+0.4; NT[i++]=-0.6; //  Abbott et al.

H0[i]=73.24; PT[i]=+1.74; NT[i++]=-1.74; //  Benetti et al.

H0[i]=72.5; PT[i]=+2.1; NT[i++]=-2.1; //  Bolejko

H0[i]=68.1; PT[i]=+2; NT[i++]=-2; //  Bolejko

H0[i]=76; PT[i]=+8; NT[i++]=-8; //  Braatz

H0[i]=69.3; PT[i]=+4.2; NT[i++]=-4.2; //  Braatz

H0[i]=71.9; PT[i]=+7.1; NT[i++]=-7.1; //  Cantiello et al.

H0[i]=73.24; PT[i]=+1.74; NT[i++]=-1.74; //  Chen

H0[i]=67.4; PT[i]=+0.5; NT[i++]=-0.5; //  Chen

H0[i]=73.24; PT[i]=+1.74; NT[i++]=-1.74; //  Choudhury & Choubey

H0[i]=72.8; PT[i]=+1.6; NT[i++]=-1.6; //  Dhawan et al.

H0[i]=55; PT[i]=+7; NT[i++]=-20; //  Di Valentino & Melchiorri

H0[i]=67.06; PT[i]=+1.68; NT[i++]=-1.68; //  Gomez-Valent

H0[i]=68.9; PT[i]=+1.96; NT[i++]=-1.96; //  Gomez-Valent

H0[i]=68.45; PT[i]=+2; NT[i++]=-2; //  Gomez-Valent

H0[i]=73.5; PT[i]=+4.6; NT[i++]=-4.7; //  Grillo

H0[i]=72.8; PT[i]=+4.3; NT[i++]=-4.1; //  Grillo

H0[i]=69.8; PT[i]=+5.3; NT[i++]=-4.1; //  Grillo

H0[i]=70.38; PT[i]=+0.6; NT[i++]=-0.6; //  Hoeneisen et al.

H0[i]=71.17; PT[i]=+1.66; NT[i++]=-1.66; //  Lee & Jang

H0[i]=73.52; PT[i]=+1.62; NT[i++]=-1.62; //  Riess et al.

H0[i]=73.83; PT[i]=+1.48; NT[i++]=-1.48; //  Riess et al.

H0[i]=73.48; PT[i]=+1.66; NT[i++]=-1.66; //  Riess et al.

H0[i]=74.4; PT[i]=+4.9; NT[i++]=-4.9; //  Van Putten

H0[i]=74.5; PT[i]=+7.3; NT[i++]=-7.3; //  Van Putten

H0[i]=74.9; PT[i]=+2.6; NT[i++]=-2.6; //  Van Putten

H0[i]=66.8; PT[i]=+1.9; NT[i++]=-1.9; //  Van Putten

H0[i]=73.75; PT[i]=+1.44; NT[i++]=-1.44; //  Van Putten

H0[i]=70; PT[i]=+12; NT[i++]=-8; //  Vitale

H0[i]=67; PT[i]=+4; NT[i++]=-4; //  Yu et al.

H0[i]=67.498; PT[i]=+7.97; NT[i++]=-3.278; //  Zhang

H0[i]=70; PT[i]=+12; NT[i++]=-8; //  Abbott

H0[i]=72.5; PT[i]=+2.5; NT[i++]=-8; //  Bethapudi & Desai

H0[i]=71.9; PT[i]=+2.4; NT[i++]=-3; //  Bonvin et al.

H0[i]=69.2; PT[i]=+1.4; NT[i++]=-2.2; //  Bonvin et al.

H0[i]=79; PT[i]=+4.4; NT[i++]=-4.2; //  Bonvin et al.

H0[i]=73.75; PT[i]=+2.11; NT[i++]=-2.11; //  Cardona

H0[i]=67.81; PT[i]=+0.92; NT[i++]=-0.92; //  Cardona

H0[i]=66.93; PT[i]=+0.62; NT[i++]=-0.62; //  Cardona

H0[i]=73.46; PT[i]=+1.4; NT[i++]=-1.4; //  Cardona

H0[i]=68.3; PT[i]=+2.7; NT[i++]=-2.6; //  Chen Yun et al.

H0[i]=68.4; PT[i]=+2.9; NT[i++]=-3.3; //  Chen Yun et al.

H0[i]=65; PT[i]=+6.6; NT[i++]=-6.6; //  Chen Yun et al.

H0[i]=67.9; PT[i]=+2.4; NT[i++]=-2.4; //  Chen Yun et al.

H0[i]=68; PT[i]=+2.8; NT[i++]=-2.8; //  Farooq

H0[i]=73.24; PT[i]=+1.74; NT[i++]=-1.74; //  Farooq

H0[i]=72.72; PT[i]=+1.67; NT[i++]=-1.67; //  Feeney et al.

H0[i]=73.15; PT[i]=+1.78; NT[i++]=-1.78; //  Feeney et al.

H0[i]=67.6; PT[i]=+0.7; NT[i++]=-0.6; //  Grieb et al.

H0[i]=73; PT[i]=+1.75; NT[i++]=-1.75; //  Guo & Zhang

H0[i]=73.24; PT[i]=+1.74; NT[i++]=-1.74; //  Hjorth et al.

H0[i]=69.13; PT[i]=+0.24; NT[i++]=-0.24; //  Huang and Huang

H0[i]=71.66; PT[i]=+1.8; NT[i++]=-1.8; //  Jang & Lee

H0[i]=73.72; PT[i]=+2.03; NT[i++]=-2.03; //  Jang & Lee

H0[i]=71.17; PT[i]=+1.66; NT[i++]=-1.66; //  Jang & Lee

H0[i]=66.2; PT[i]=+8.9; NT[i++]=-8.9; //  Pritychenko

H0[i]=67.2; PT[i]=+6.9; NT[i++]=-6.9; //  Pritychenko

H0[i]=69.13; PT[i]=+2.34; NT[i++]=-2.34; //  Wang et al.

H0[i]=73.24; PT[i]=+1.74; NT[i++]=-1.74; //  Wei & Wu

H0[i]=69.6; PT[i]=+0.7; NT[i++]=-0.7; //  Wei & Wu

H0[i]=73.1; PT[i]=+5.7; NT[i++]=-6; //  Wong et al.

H0[i]=72.5; PT[i]=+3.1; NT[i++]=-3.1; //  Zhang et al.

H0[i]=67.8; PT[i]=+0.9; NT[i++]=-0.9; //  Ade et al.

H0[i]=66; PT[i]=+6; NT[i++]=-6; //  Gao et al.

H0[i]=70.1; PT[i]=+0.34; NT[i++]=-0.34; //  Ichiki et al.

H0[i]=66.5; PT[i]=+1.8; NT[i++]=-1.8; //  Ludovic et al.

H0[i]=64.2; PT[i]=+1.9; NT[i++]=-1.9; //  Ludovic et al.

H0[i]=91.8; PT[i]=+5.3; NT[i++]=-5.3; //  Moresco et al.

H0[i]=72.25; PT[i]=+2.51; NT[i++]=-2.51; //  Riess et al.

H0[i]=72.04; PT[i]=+2.67; NT[i++]=-2.67; //  Riess et al.

H0[i]=76.18; PT[i]=+2.37; NT[i++]=-2.37; //  Riess et al.

H0[i]=74.5; PT[i]=+3.27; NT[i++]=-3.27; //  Riess et al.

H0[i]=73.24; PT[i]=+1.74; NT[i++]=-1.74; //  Riess et al.

H0[i]=76.2; PT[i]=+3.4; NT[i++]=-3.4; //  Tully et al.

H0[i]=75; PT[i]=+2; NT[i++]=-2; //  Tully et al.

H0[i]=68.17; PT[i]=+1.55; NT[i++]=-1.56; //  Cheng & Qing Guo

H0[i]=68.11; PT[i]=+1.69; NT[i++]=-1.69; //  Cheng & Qing Guo

H0[i]=68.11; PT[i]=+0.86; NT[i++]=-0.86; //  Cheng & Qing Guo

H0[i]=67.7; PT[i]=+1.1; NT[i++]=-1.1; //  Cuesta et al.

H0[i]=69.8; PT[i]=+2.6; NT[i++]=-2.6; //  Jang & Lee

H0[i]=72.2; PT[i]=+3.3; NT[i++]=-3.3; //  Jang & Lee

H0[i]=68.1; PT[i]=+5.9; NT[i++]=-5.9; //  Kumar et al.

H0[i]=73; PT[i]=+26; NT[i++]=-22; //  Kuo et al.

H0[i]=70.6; PT[i]=+2.6; NT[i++]=-2.6; //  Rigault et al.

H0[i]=68.8; PT[i]=+3.3; NT[i++]=-3.3; //  Rigault et al.

H0[i]=67.3; PT[i]=+1.2; NT[i++]=-1.2; //  Ade et al.

H0[i]=70.8; PT[i]=+2.4; NT[i++]=-2.4; //  Ben-Dayan et al.

H0[i]=69.6; PT[i]=+0.7; NT[i++]=-0.7; //  Bennett et al.

H0[i]=64.9; PT[i]=+4.2; NT[i++]=-4.2; //  Busti et al.

H0[i]=72.5; PT[i]=+2.5; NT[i++]=-2.5; //  Efstathiou

H0[i]=70.6; PT[i]=+3.3; NT[i++]=-3.3; //  Efstathiou

H0[i]=74.1; PT[i]=+2.2; NT[i++]=-2.2; //  Lima & Cunha

H0[i]=70; PT[i]=+2.2; NT[i++]=-2.2; //  Bennett et al.

H0[i]=69.32; PT[i]=+0.8; NT[i++]=-0.8; //  Bennett et al.

H0[i]=68; PT[i]=+4.8; NT[i++]=-4.8; //  Braatz et al.

H0[i]=68; PT[i]=+2.8; NT[i++]=-2.8; //  Farooq & Bathra

H0[i]=73.8; PT[i]=+2.4; NT[i++]=-2.4; //  Farooq & Bathra

H0[i]=69.7; PT[i]=+2.4; NT[i++]=-2.4; //  Hinshaw et al.

H0[i]=70.4; PT[i]=+2.5; NT[i++]=-2.5; //  Hinshaw et al.

H0[i]=69.33; PT[i]=+0.88; NT[i++]=-0.88; //  Hinshaw et al.

H0[i]=70.2; PT[i]=+1.4; NT[i++]=-1.4; //  Hinshaw et al.

H0[i]=70; PT[i]=+3; NT[i++]=-3; //  Humphreys et al.

H0[i]=68; PT[i]=+9; NT[i++]=-9; //  Kuo et al.

H0[i]=49.97; PT[i]=+0.19; NT[i++]=-0.19; //  Pietrzynski et al.

H0[i]=68.9; PT[i]=+7.1; NT[i++]=-7.1; //  Reid et al.

H0[i]=72.1; PT[i]=+3.2; NT[i++]=-2.3; //  Salvatelli et al.

H0[i]=74.1; PT[i]=+2.1; NT[i++]=-2.1; //  Scowcroft et al.

H0[i]=69; PT[i]=+6; NT[i++]=-6; //  Sereno and Pacificz

H0[i]=80; PT[i]=+5.8; NT[i++]=-5.7; //  Suyu et al.

H0[i]=75.2; PT[i]=+4.4; NT[i++]=-4.2; //  Suyu et al.

H0[i]=73.1; PT[i]=+2.4; NT[i++]=-3.6; //  Suyu et al.

H0[i]=74.4; PT[i]=+3; NT[i++]=-3; //  Tully et al.

H0[i]=71.3; PT[i]=+2; NT[i++]=-2; //  Xia et al.

H0[i]=73.8; PT[i]=+2.4; NT[i++]=-2.4; //  Calabrese et al.

H0[i]=68; PT[i]=+2.8; NT[i++]=-2.8; //  Calabrese et al.

H0[i]=69.7; PT[i]=+2.5; NT[i++]=-2.5; //  Calabrese et al.

H0[i]=74.3; PT[i]=+3.1; NT[i++]=-3.1; //  Chavez

H0[i]=67; PT[i]=+3.2; NT[i++]=-3.2; //  Colless et al.

H0[i]=74.3; PT[i]=+3; NT[i++]=-3; //  Freedman et al.

H0[i]=70.2; PT[i]=+0.14; NT[i++]=-0.14; //  Pozzo

H0[i]=75.4; PT[i]=+2.9; NT[i++]=-2.9; //  Riess et al.

H0[i]=56; PT[i]=+2; NT[i++]=-2; //  Wang

H0[i]=68; PT[i]=+5.5; NT[i++]=-5.5; //  Chen & Ratra

H0[i]=67; PT[i]=+3.2; NT[i++]=-3.2; //  Beutler et al.

H0[i]=71; PT[i]=+2.5; NT[i++]=-2.5; //  Jarosik et al.

H0[i]=70.4; PT[i]=+1.3; NT[i++]=-1.4; //  Jarosik et al.

H0[i]=74.8; PT[i]=+3.1; NT[i++]=-3.1; //  Riess et al.

H0[i]=74.4; PT[i]=+2.5; NT[i++]=-2.5; //  Riess et al.

H0[i]=73.8; PT[i]=+2.4; NT[i++]=-2.4; //  Riess et al.

H0[i]=73; PT[i]=+2; NT[i++]=-2; //  Freedman & Madore

H0[i]=66; PT[i]=+6; NT[i++]=-4; //  Paraficz et Hjorth

H0[i]=76; PT[i]=+3; NT[i++]=-3; //  Paraficz et Hjorth

H0[i]=70.6; PT[i]=+3.1; NT[i++]=-3.1; //  Suyu et al.

H0[i]=69.7; PT[i]=+4.9; NT[i++]=-5; //  Suyu et al.

H0[i]=70.5; PT[i]=+1.3; NT[i++]=-1.3; //  Hinshaw et al.

H0[i]=71.9; PT[i]=+2.6; NT[i++]=-2.7; //  Hinshaw et al.

H0[i]=70.5; PT[i]=+1.3; NT[i++]=-1.3; //  Komatsu et al.

H0[i]=70.4; PT[i]=+1.4; NT[i++]=-1.4; //  Komatsu et al.

H0[i]=70.9; PT[i]=+1.3; NT[i++]=-1.3; //  Komatsu et al.

H0[i]=70.1; PT[i]=+1.3; NT[i++]=-1.3; //  Komatsu et al.

H0[i]=74.2; PT[i]=+3.6; NT[i++]=-3.6; //  Riess et al.

H0[i]=84.2; PT[i]=+6; NT[i++]=-6; //  Russell

H0[i]=83.4; PT[i]=+8; NT[i++]=-8; //  Russell

H0[i]=88; PT[i]=+6; NT[i++]=-6; //  Russell

H0[i]=61.7; PT[i]=+1.2; NT[i++]=-1.1; //  Leith et al.

H0[i]=67; PT[i]=+13; NT[i++]=-10; //  Vuissoz et al.

H0[i]=63; PT[i]=+7; NT[i++]=-3; //  Vuissoz et al.

H0[i]=70; PT[i]=+6; NT[i++]=-6; //  Oguri

H0[i]=68; PT[i]=+6; NT[i++]=-6; //  Oguri

H0[i]=73.5; PT[i]=+3.2; NT[i++]=-3.2; //  Spergel et al.

H0[i]=73.2; PT[i]=+3.1; NT[i++]=-3.2; //  Spergel et al.

H0[i]=70.4; PT[i]=+1.5; NT[i++]=-1.6; //  Spergel et al.

H0[i]=76.9; PT[i]=+3.9; NT[i++]=-3.4; //  Bonamente et al.

H0[i]=73.7; PT[i]=+4.6; NT[i++]=-3.8; //  Bonamente et al.

H0[i]=77.6; PT[i]=+4.8; NT[i++]=-4.3; //  Bonamente et al.

H0[i]=70.8; PT[i]=+1.9; NT[i++]=-1.8; //  Hütsi

H0[i]=74.92; PT[i]=+2.28; NT[i++]=-2.28; //  Ngeow and Kanbur

H0[i]=74.37; PT[i]=+2.27; NT[i++]=-2.27; //  Ngeow and Kanbur

H0[i]=62.3; PT[i]=+1.3; NT[i++]=-1.3; //  Sandage et al.

H0[i]=60.9; PT[i]=+1.3; NT[i++]=-1.3; //  Sandage et al.

H0[i]=60.7; PT[i]=+1.5; NT[i++]=-1.5; //  Sandage et al.

H0[i]=72; PT[i]=+6; NT[i++]=-6; //  Wang et al.

H0[i]=73.2; PT[i]=+7; NT[i++]=-7; //  Gibson & Brook

H0[i]=75; PT[i]=+7; NT[i++]=-7; //  Hamuy

H0[i]=65; PT[i]=+12; NT[i++]=-12; //  Hamuy

H0[i]=58; PT[i]=+2; NT[i++]=-2; //  Magain

H0[i]=58; PT[i]=+2; NT[i++]=-2; //  Olivares et al.

H0[i]=73; PT[i]=+4; NT[i++]=-4; //  Riess

H0[i]=69; PT[i]=+8; NT[i++]=-8; //  Schmidt et al.

H0[i]=66; PT[i]=+8; NT[i++]=-8; //  Stritzinger et al.

H0[i]=78; PT[i]=+9; NT[i++]=-9; //  Stritzinger et al.

H0[i]=67; PT[i]=+30; NT[i++]=-18; //  Udomprasert et al.

H0[i]=64; PT[i]=+7; NT[i++]=-4; //  Boffi & Riess

H0[i]=33; PT[i]=+5; NT[i++]=-5; //  Dumin

H0[i]=69; PT[i]=+12; NT[i++]=-12; //  Jimenez et al.

H0[i]=75; PT[i]=+7; NT[i++]=-6; //  Koopmans

H0[i]=70; PT[i]=+7; NT[i++]=-7; //  Mei et al.

H0[i]=68; PT[i]=+6; NT[i++]=-6; //  Mei et al.

H0[i]=68; PT[i]=+5; NT[i++]=-5; //  Mei et al.

H0[i]=71; PT[i]=+4; NT[i++]=-4; //  Mei et al.

H0[i]=77; PT[i]=+19; NT[i++]=-15; //  Saunders et al.

H0[i]=85; PT[i]=+20; NT[i++]=-17; //  Saunders et al.

H0[i]=72; PT[i]=+5; NT[i++]=-5; //  Spergel et al.

H0[i]=71; PT[i]=+4; NT[i++]=-3; //  Spergel et al.

H0[i]=63; PT[i]=+2; NT[i++]=-2; //  Fassnacht et al.

H0[i]=72; PT[i]=+8; NT[i++]=-8; //  Freedman

H0[i]=57; PT[i]=+23; NT[i++]=-16; //  Grainge et al.

H0[i]=48; PT[i]=+7; NT[i++]=-4; //  Kochanek

H0[i]=71; PT[i]=+6; NT[i++]=-6; //  Kochanek

H0[i]=72; PT[i]=+8; NT[i++]=-8; //  Kochanek

H0[i]=62; PT[i]=+7; NT[i++]=-7; //  Kochanek

H0[i]=75; PT[i]=+8; NT[i++]=-8; //  Tikhonov & Galazoutdinova

H0[i]=81; PT[i]=+5; NT[i++]=-5; //  Tikhonov & Galazoutdinova

H0[i]=59; PT[i]=+15; NT[i++]=-10; //  Treu & Koopmans

H0[i]=71; PT[i]=+2; NT[i++]=-2; //  Freedman et al.

H0[i]=71; PT[i]=+3; NT[i++]=-3; //  Freedman et al.

H0[i]=70; PT[i]=+5; NT[i++]=-5; //  Freedman et al.

H0[i]=72; PT[i]=+9; NT[i++]=-9; //  Freedman et al.

H0[i]=82; PT[i]=+6; NT[i++]=-6; //  Freedman et al.

H0[i]=72; PT[i]=+8; NT[i++]=-8; //  Freedman et al.

H0[i]=65; PT[i]=+5; NT[i++]=-5; //  Itoh

H0[i]=76; PT[i]=+1.3; NT[i++]=-1.3; //  Jensen et al.

H0[i]=72; PT[i]=+2.3; NT[i++]=-2.3; //  Jensen et al.

H0[i]=65; PT[i]=+5; NT[i++]=-5; //  Koopmans et al.

H0[i]=71; PT[i]=+8; NT[i++]=-8; //  Liu & Graham

H0[i]=64; PT[i]=+14; NT[i++]=-18; //  Mason et al.

H0[i]=66; PT[i]=+14; NT[i++]=-11; //  Mason et al.

H0[i]=70; PT[i]=+7; NT[i++]=-7; //  Mei et al.

H0[i]=69; PT[i]=+4; NT[i++]=-4; //  Tonry

H0[i]=71; PT[i]=+6; NT[i++]=-6; //  Willick & Puneet

H0[i]=63; PT[i]=+4.3; NT[i++]=-4.3; //  Xiao-Feng et al.

H0[i]=69; PT[i]=+4; NT[i++]=-4; //  Ferrarese et al.

H0[i]=68; PT[i]=+2; NT[i++]=-2; //  Gibson et al.

H0[i]=71; PT[i]=+6; NT[i++]=-6; //  Mould et al.

H0[i]=68; PT[i]=+6; NT[i++]=-6; //  Mould et al.

H0[i]=71; PT[i]=+4; NT[i++]=-4; //  Sakai et al.

H0[i]=77; PT[i]=+7; NT[i++]=-7; //  Tikhonov et al.

H0[i]=69; PT[i]=+12; NT[i++]=-19; //  Biggs et al.

H0[i]=69; PT[i]=+18; NT[i++]=-12; //  Chae KH

H0[i]=74; PT[i]=+18; NT[i++]=-17; //  Chae KH

H0[i]=42; PT[i]=+9; NT[i++]=-9; //  Collier et al.

H0[i]=73; PT[i]=+6; NT[i++]=-6; //  Freedman et al.

H0[i]=64; PT[i]=+8; NT[i++]=-6; //  Jha et al.

H0[i]=85; PT[i]=+27; NT[i++]=-23; //  Mason & Myers

H0[i]=61; PT[i]=+15; NT[i++]=-14; //  Mason & Myers

H0[i]=61; PT[i]=+23; NT[i++]=-21; //  Mason & Myers

H0[i]=80; PT[i]=+19; NT[i++]=-17; //  Mason & Myers

H0[i]=68; PT[i]=+21; NT[i++]=-19; //  Mason & Myers

H0[i]=71; PT[i]=+5; NT[i++]=-5; //  Mason & Myers

H0[i]=86; PT[i]=+24; NT[i++]=-24; //  Mazumdar & Narasimba

H0[i]=67; PT[i]=+7; NT[i++]=-7; //  Tanvir et al.

H0[i]=62.9; PT[i]=+1.6; NT[i++]=-1.6; //  Tripp & Branch

H0[i]=62; PT[i]=+2; NT[i++]=-2; //  Tripp & Branch

H0[i]=60; PT[i]=+10; NT[i++]=-10; //  Branch

H0[i]=66; PT[i]=+15; NT[i++]=-14; //  Goicoechea et al.

H0[i]=77; PT[i]=+8; NT[i++]=-8; //  Harris et al.

H0[i]=47; PT[i]=+23; NT[i++]=-15; //  Hughes & Birkinshaw

H0[i]=82; PT[i]=+8; NT[i++]=-8; //  Lauer et al.

H0[i]=89; PT[i]=+10; NT[i++]=-10; //  Lauer et al.

H0[i]=65.2; PT[i]=+1.3; NT[i++]=-1.3; //  Riess et al.

H0[i]=63.8; PT[i]=+1.3; NT[i++]=-1.3; //  Riess et al.

H0[i]=55; PT[i]=+8; NT[i++]=-8; //  Tammann & Labhardt

H0[i]=60; PT[i]=+6; NT[i++]=-6; //  Tripp

H0[i]=70; PT[i]=+5; NT[i++]=-5; //  Giovanelli

H0[i]=76; PT[i]=+8; NT[i++]=-8; //  Giovanelli

H0[i]=67; PT[i]=+8; NT[i++]=-8; //  Giovanelli

H0[i]=75; PT[i]=+6; NT[i++]=-6; //  Gregg

H0[i]=67; PT[i]=+8; NT[i++]=-8; //  Hjorth & Tanvir

H0[i]=70; PT[i]=+7; NT[i++]=-7; //  Hjorth & Tanvir

H0[i]=60; PT[i]=+40; NT[i++]=-23; //  Holzapfel et al.

H0[i]=78; PT[i]=+34; NT[i++]=-28; //  Holzapfel et al.

H0[i]=78; PT[i]=+60; NT[i++]=-40; //  Holzapfel et al.

H0[i]=58; PT[i]=+10; NT[i++]=-5; //  Hoyle et al.

H0[i]=74; PT[i]=+10; NT[i++]=-10; //  Schechter

H0[i]=52.5; PT[i]=+2.5; NT[i++]=-2.5; //  Sciama

H0[i]=54.8; PT[i]=+0.3; NT[i++]=-0.3; //  Sciama

H0[i]=81; PT[i]=+6; NT[i++]=-6; //  Tonry et al.

H0[i]=69; PT[i]=+8; NT[i++]=-8; //  Amendola

H0[i]=80; PT[i]=+17; NT[i++]=-17; //  Amendola

H0[i]=49.5; PT[i]=+4.5; NT[i++]=-4.5; //  Biesiada

H0[i]=65; PT[i]=+8; NT[i++]=-8; //  Forbes et al.

H0[i]=103; PT[i]=+59; NT[i++]=-28; //  Kobayashi

H0[i]=82; PT[i]=+56; NT[i++]=-24; //  Kobayashi

H0[i]=60; PT[i]=+24; NT[i++]=-13; //  Kobayashi

H0[i]=51; PT[i]=+10; NT[i++]=-7; //  Kobayashi

H0[i]=33; PT[i]=+22; NT[i++]=-9; //  Kobayashi

H0[i]=74; PT[i]=+26; NT[i++]=-15; //  Kobayashi

H0[i]=63; PT[i]=+28; NT[i++]=-15; //  Kobayashi

H0[i]=80; PT[i]=+17; NT[i++]=-17; //  Mallik

H0[i]=87; PT[i]=+7; NT[i++]=-7; //  Mallik

H0[i]=55; PT[i]=+3; NT[i++]=-3; //  Schaefer

H0[i]=56; PT[i]=+3; NT[i++]=-3; //  Schaefer

H0[i]=82.5; PT[i]=+5.9; NT[i++]=-3; //  Grogin & Narayan

H0[i]=82.5; PT[i]=+8.7; NT[i++]=-5.6; //  Grogin & Narayan

H0[i]=71; PT[i]=+30; NT[i++]=-25; //  Herbig

H0[i]=74.6; PT[i]=+47; NT[i++]=-33; //  Holzapfel et al.

H0[i]=38; PT[i]=+18; NT[i++]=-16; //  Jones

H0[i]=80; PT[i]=+17; NT[i++]=-17; //  Kennicutt Jr et al.

H0[i]=73; PT[i]=+11; NT[i++]=-11; //  Mould

H0[i]=81; PT[i]=+11; NT[i++]=-11; //  Mould

H0[i]=84; PT[i]=+16; NT[i++]=-16; //  Mould

H0[i]=76; PT[i]=+10; NT[i++]=-10; //  Mould

H0[i]=82; PT[i]=+11; NT[i++]=-11; //  Mould

H0[i]=71; PT[i]=+10; NT[i++]=-10; //  Mould

H0[i]=80; PT[i]=+17; NT[i++]=-17; //  Mould

H0[i]=80; PT[i]=+17; NT[i++]=-17; //  Nakamura & Suto

H0[i]=58; PT[i]=+6; NT[i++]=-6; //  Rephaeli

H0[i]=51; PT[i]=+7; NT[i++]=-7; //  Schaefer

H0[i]=61; PT[i]=+12; NT[i++]=-12; //  Schaefer

H0[i]=26; PT[i]=+5; NT[i++]=-5; //  Schaefer

H0[i]=69; PT[i]=+8; NT[i++]=-8; //  Tanvir et al.

H0[i]=78; PT[i]=+11; NT[i++]=-11; //  Whitmore & Schweizer

H0[i]=65; PT[i]=+25; NT[i++]=-25; //  Birkinshaw & Hughes

H0[i]=55; PT[i]=+17; NT[i++]=-17; //  Birkinshaw & Hughes

H0[i]=80; PT[i]=+17; NT[i++]=-17; //  Freedman

H0[i]=84; PT[i]=+5; NT[i++]=-5; //  Lu et al.

H0[i]=73; PT[i]=+6; NT[i++]=-6; //  Schmidt & Kirshner

H0[i]=90; PT[i]=+10; NT[i++]=-10; //  Tully

H0[i]=43.5; PT[i]=+2.7; NT[i++]=-2.7; //  Duemmler

H0[i]=77; PT[i]=+8; NT[i++]=-8; //  Lauer & Postman

H0[i]=51; PT[i]=+5; NT[i++]=-5; //  Lauer & Postman

H0[i]=75; PT[i]=+30; NT[i++]=-30; //  Leibundgut & Pinto

H0[i]=40; PT[i]=+9; NT[i++]=-9; //  Birkinshaw

H0[i]=45; PT[i]=+12; NT[i++]=-12; //  Birkinshaw

H0[i]=82; PT[i]=+7; NT[i++]=-7; //  Tonry

H0[i]=52; PT[i]=+2; NT[i++]=-2; //  Sandage & Tammann

H0[i]=45; PT[i]=+3; NT[i++]=-3; //  Sandage & Tammann

H0[i]=73; PT[i]=+10; NT[i++]=-10; //  Visvanathan

H0[i]=50; PT[i]=+10; NT[i++]=-10; //  Sandage & Tammann

H0[i]=52; PT[i]=+2; NT[i++]=-2; //  Sandage & Tammann

H0[i]=50; PT[i]=+7; NT[i++]=-7; //  Sandage & Tammann

H0[i]=67; PT[i]=+10; NT[i++]=-10; //  Dressler

H0[i]=74.3; PT[i]=+11; NT[i++]=-11; //  Visvanathan

H0[i]=74.3; PT[i]=+11; NT[i++]=-11; //  Hanes

H0[i]=76; PT[i]=+8; NT[i++]=-8; //  Bottinelli & Gouguenheim

H0[i]=50.3; PT[i]=+4.3; NT[i++]=-4.3; //  Sandage & Tammann

H0[i]=56.9; PT[i]=+3.4; NT[i++]=-3.4; //  Sandage & Tammann

H0[i]=57; PT[i]=+6; NT[i++]=-6; //  Sandage & Tammann

H0[i]=55.5; PT[i]=+8.7; NT[i]=-8.7; //  Sandage & Tammann

//Creates an H0 array that contains all the extremities of the tolerance ranges

for (i=0;i<=nbH0-1;i++) {

H0[i+nbH0]=H0[i]+PT[i]; Variation[i+nbH0]=PT[i];

H0[i]=H0[i]+NT[i]; Variation[i]=NT[i]; }

//Sorts H0 array in ascending order with corresponding Variation of tolerance

for (j=0;j<=2*nbH0-2;j++) {

for (i=j+1;i<=2*nbH0-1;i++) {

if (H0[i]

DH0=H0[j]; DSgn=Variation[j]; H0[j]=H0[i];

Variation[j]=Variation[i]; H0[i]=DH0; Variation[i]=DSgn; } } }

for (i=0;i<=2*nbH0-1;i++) { //Builds the nbCrossings array

if (i==0) { nbCrossings[i]=1; }

else {

if (Variation[i]<0) {nbCrossings[i]=nbCrossings[i-1]+1;}

if (Variation[i]>0) {nbCrossings[i]=nbCrossings[i-1]-1;}

if (H0[i]==H0[i-1]) {

j=i; DeltaCrossings=0;

do {

if (Variation[j]<0) {DeltaCrossings=DeltaCrossings+1;}

if (Variation[j]>0) {DeltaCrossings=DeltaCrossings-1;}

j=j-1; } while (H0[j]==H0[i]);

for(k=i;k>j;k--) {nbCrossings[k]=nbCrossings[j]+DeltaCrossings;}

} } } } //End of CreateTableOfCrossingH0Ranges

//***Function that returns the y coordinate corresponding to x for non

//***centered Gaussian curve

double GaussianCurve(double x, double Mean, double Sigma, double Multiplier) {

double y; //Coordinate y corresponding to x for a non centered Gaussian

y=(Multiplier/(Sigma*Sqrt(2*Pi)))*exp(-0.5*Sqr((x-Mean)/Sigma));

return y; } //End of CreateApproximativeCurve

//***This the best Gaussians to fit the nbCrossing array as a function of H0***

double FindsGaussianCurvesLS(double Mean,double Sigma, double Multiplier) {

int i,j; /*Counters*/ double LS = 0; //Least square

double Sum; //Sum of the 5 Gaussien curve for a specific H0 value

for (j=0;j<=2*nbH0-1;j++) {

Sum=0;

for (i=0;i<=4;i++) {

Sum=Sum+GaussianCurve(H0[j],Mean[i],Sigma[i],Multiplier[i]); }

//We give a heavier weight to any error between 69.2 and 72.1 to

// model the gap between these values

if ((H0[i]>=69.2)&&(H0[i]<=72.1)) {

LS=LS+10*(Sqr(nbCrossings[j]-Sum));}

else { LS=LS+Sqr(nbCrossings[j]-Sum); }

} return LS; } //End of FindsGaussianCurvesLS

//***This function finds the best Gaussians to fit the real curve

void FindsBestGaussiansToFitRealCurve(void) {

int i; //Counter

double DMean,DSigma,DMultiplier; //Dummy arrays

double LS, PLS; //Least Square and Previous Least Square

double StepMean = 0.1, StepSigma = 0.1, StepMultiplier = 0.1; //Variations

int nbMeanNotImproved = 0; //Tells how many times not improved

int nbSigmaNotImproved = 0; //Tells how many times not improved

int nbMultiplierNotImproved = 0; //Tells how many times not improved

//Starting values (approximative values only)

Mean=71; Sigma=1; Multiplier=-280;

Mean=68; Sigma=17; Multiplier=3800;

Mean=Mean; Sigma=3; Multiplier=880;

Mean=73; Sigma=7; Multiplier=1200;

Mean=Mean; Sigma=2; Multiplier=470;

//Fills the 3 dummy arrays DMean, DSigma and DMultiplier

//with the same values than the arrays Mean, Sigma and Multiplier

for(i=0;i<=4;i++) {

DMean[i]=Mean[i]; DSigma[i]=Sigma[i]; DMultiplier[i]=Multiplier[i]; }

//Tries to find the 5 best Gaussians to fit the curve

do {

for (i=0;i<=4;i++) {

//We improve Mean[i], but we force

//Mean = Mean & Mean = Mean

if ((i!=2)&&(i!=4)) {

PLS=FindsGaussianCurvesLS(Mean,Sigma,Multiplier);

DMean[i]=Mean[i]+StepMean;

if (i==1) {DMean=DMean[i];}

if (i==3) {DMean=DMean[i];}

LS=FindsGaussianCurvesLS(DMean,DSigma,DMultiplier);

if (LS

Mean[i]=DMean[i];

if (i==1) {Mean=DMean[i];}

if (i==3) {Mean=DMean[i];}

nbMeanNotImproved=0; }

else {

DMean[i]=Mean[i]-StepMean;

if (i==1) {DMean=DMean[i];}

if (i==3) {DMean=DMean[i];}

LS=FindsGaussianCurvesLS(DMean,DSigma,DMultiplier);

if (LS

Mean[i]=DMean[i];

if (i==1) {Mean=DMean[i];}

if (i==3) {Mean=DMean[i];}

nbMeanNotImproved=0; }

else {

DMean[i]=Mean[i];

if (i==1) {DMean=DMean[i];}

if (i==3) {DMean=DMean[i];}

nbMeanNotImproved++;

if (nbMeanNotImproved>=100) {

nbMeanNotImproved=0; StepMean=StepMean/10; } } } }

//We try to improve Sigma[i]

PLS=FindsGaussianCurvesLS(Mean,Sigma,Multiplier);

DSigma[i]=Sigma[i]+StepSigma;

LS=FindsGaussianCurvesLS(DMean,DSigma,DMultiplier);

if (LS

else {

DSigma[i]=Sigma[i]-StepSigma;

LS=FindsGaussianCurvesLS(DMean,DSigma,DMultiplier);

if (LS

Sigma[i]=DSigma[i]; nbSigmaNotImproved=0; }

else {

DSigma[i]=Sigma[i]; nbSigmaNotImproved++;

if (nbSigmaNotImproved>=100) {

nbSigmaNotImproved=0; StepSigma=StepSigma/10; } } }

//We try to improve Multiplier[i]

PLS=FindsGaussianCurvesLS(Mean,Sigma,Multiplier);

DMultiplier[i]=Multiplier[i]+StepMultiplier;

LS=FindsGaussianCurvesLS(DMean,DSigma,DMultiplier);

if (LS

Multiplier[i]=DMultiplier[i]; nbMultiplierNotImproved=0; }

else {

DMultiplier[i]=Multiplier[i]-StepMultiplier;

LS=FindsGaussianCurvesLS(DMean,DSigma,DMultiplier);

if (LS

Multiplier[i]=DMultiplier[i]; nbMultiplierNotImproved=0; }

else {

DMultiplier[i]=Multiplier[i]; nbMultiplierNotImproved++;

if (nbMultiplierNotImproved>=100) {

nbMultiplierNotImproved=0;

StepMultiplier=StepMultiplier/10; } } } }

LS=FindsGaussianCurvesLS(DMean,DSigma,DMultiplier);

} while (LS>=22000); //Sets a stop point

for(i=0;i<=4;i++) {

printf("\n Mean[%i]=%10lf Sigma[%i]=%10lf Multiplier[%i]=%10lf",

i,Mean[i],i,Sigma[i],i,Multiplier[i]);

} } //End of FindsBestGaussiansToFitRealCurve

//***The function begins by shifting Mean and Mean of the two

//***Gaussian curves that are around H0=69.2km/(s*MParsec) to

//***H0=72.1km/(s*MParsec) with a theoretical factor of 1.042516951. Then

//***the function that modifies the global H0 array builds the summation of

//***the 4 positive Gaussian curves.

void CreatesFinalGaussianCurve(void) {

int i; //Dummy index value

//We shift Mean and Mean with a theoretical factor of 1.042516951

Mean=Mean*1.042516951; Mean=Mean;

//We omit i=0 to remove the negative Gaussian curve

for(i=1;i<=nbH0*2-1;i++) {

nbCrossings[i]=GaussianCurve(H0[i],Mean,Sigma,Multiplier);

nbCrossings[i]=nbCrossings[i]+GaussianCurve(H0[i],Mean,Sigma,

Multiplier);

nbCrossings[i]=nbCrossings[i]+GaussianCurve(H0[i],Mean,Sigma,

Multiplier);

nbCrossings[i]=nbCrossings[i]+GaussianCurve(H0[i],Mean,Sigma,

Multiplier); } } //End of CreatesFinalGaussianCurve

//***Returns the Least Mean Square of the equation DA*x^2+DB*x+DC.

double FindsLMS(double DA, double DB, double DC) {

int i; /*Dummy index value*/ double LMS = 0; //Least Mean Square

for(i=PosTipIndex-n;i<=PosTipIndex+n;i++){

LMS=LMS+Sqr(nbCrossings[i]-(DA*H0[i]*H0[i]+DB*H0[i]+DC));

} return LMS; } //End of FindsLMS

//***Returns A, B, and C coefficients of the quadratic equation of the tip

void ApproximatesTipEquation(void) {

double LMSTipMin; //Reminds the lowest value of least mean square

double PLMSTip, NLMSTip; //LMSTip for a forward and backward step

double DA,DB,DC; //Dummy values of A, B and C coefficients

double StepA,StepB,StepC; //Step variation of the coefficients

double nbCrossingsMax = 0; //Maximum number crossings at the tip

double xa,xb,xc,ya,yb,yc; /*3 coordinates*/ int i; //Dummy index value

for(i=0;i<=nbH0*2-1;i++) { //Finds the index of the approximated tip

if (nbCrossings[i]>nbCrossingsMax) {

nbCrossingsMax=nbCrossings[i]; PosTipIndex=i; } }

i=PosTipIndex; xa=H0[i-n]; xb=H0[i]; xc=H0[i+n];

ya=nbCrossings[i-n]; yb=nbCrossings[i]; yc=nbCrossings[i+n];

//Sets coefficients ABC

A=((yc-ya)/((xc-xa)*(xc-xb)))-((yb-ya)/((xb-xa)*(xc-xb)));

B=((yb-ya)/(xb-xa))-A*(xb+xa); C=ya-A*xa*xa-B*xa; DA=A;DB=B;DC=C;

} //End of ApproximatesTipEquation

//***Function that returns the Best estimate of H0

void BestEstimateOfH0(void) {

CreateTableOfCrossingH0Ranges(); FindsBestGaussiansToFitRealCurve();

CreatesFinalGaussianCurve(); ApproximatesTipEquation();

BEH=-B/(2*A); Accuracy_ppm = (BEH-TVH)/(TVH*1E-6);

LMSTip=FindsLMS(A,B,C); } //End of BesEstimateOfH0

int main(void) {

BestEstimateOfH0();

printf("\n\n Equation of the tip: y = %.10lfx^2+ %.10lfx+ %.10lf",A,B,C);

printf("\n Best estimate of H0 = %.10lf km/(s*MParsec)",BEH);

printf("\n Theoretical H0 = %.10lf km/(s*MParsec)",TVH);

printf("\n Relative accuracy versus theoretical value = %.10lf ppm",

Accuracy_ppm);

getchar(); return 0;

} //End of main