Water is a conductive liquid and its condyctivity vary with the contents

of ionic impurities. Very pure water has resistivity about 15 MOhm*cm

tap water about 100 times smaller and see water about 10000 smaller.

Therfore the imaginary part of permittivity  wich depends on conductivity

as:  Im(eps)=sigma/(omega*eps0)  where omega - angular frequency

eps0 - permittivity of vacuum (about 8.854E-12 in SI units) 

predominantly  depends on comductivity at low frequencies.

At 1 MHz the imaginary part of permittivity of pure water

would be about 0.12 tap water about 12 and see water about 1200

There are a lot of papers about general properties of water 

See e.g. 

[1] Kaatze U 1997 The dielectric properties of water in its

different states of interaction J. Solut. Chem. 26 1049–112

[2] Kaatze U 2010 Techniques for measuring the microwave

dielectric properties of materials Metrologia 47 S91–113

Thanks Sir. However, I am looking for the dielectric constant and dielectric loss factor of water at radio frequency range (1 to 30 MHz).

I understand this. The real part of permittivity can be considered as a constant (frequency indepenent quantity in this frequency range) and the imaginary part has

to be calculated as

 Im(eps)=sigma/(omega*eps0)

Note that the part of the imaginary part of permittivity related to the pure

dielectric polarization mechanism of water molecules can not  be measured at

this frequency range because it is orders of magnitude smaller than the

part related to conductor losses. In any experiment the total

imaginary part of permittivity is measured being the sum of dielectric polarization

and conductor losses. 

As a reasonable starting value, you can assume a dielectric constant Epsilon_r in the range of 80. This number worked well for the water cooled RF cavities at COSY, and J-PARC in the frequency range from 0.5 MHz to 10 MHz.

Thanks Dr. Jerzy for the very informative explaination.

Thanks @Alexander for your answer. You are right, the dielectric constant of water would be around 80. I am not sure if the value of dielectric loss factor could be assumed at the range of 8. I cannot find any paper on it. 

I do not know if this will help or not but here it is:

:"Water is a rather important dielectric liquid in pulsed-power applications. It has a relatively high electric breakdown strength (up to 3 x 107 V/m) for submicrosecond electric stress and, owing to its high permittivity, can store quite large energy densities for short times. Most of the electrical characteristics of organic dielectric liquid insulators . . . are also valid for water.

A small fraction (10-7) of water molecules is always dissociated into H+ and OH-. These ions lead to a residual conductivity of 4 x 10-6 S/m even for very clean water. Therefore water is inadequate for DC-insulation. . . . Nevertheless, ionic currents do not contribute to the initiation of breakdown for submicrosecond pulses. This has been demonstrated even for salt solutions with concentrations up to 1M. . . . Water, which is largely used in short-pulse applications, has, in addition, the benefit of a high dielectric constant (e= 81), which allows one to store high energy densities.

Under short-duration electric stress, the electric strength of water becomes comparable to that of other liquid insulators. At 1 ms, its strength is around 40MV/m. . . . Its self-discharge time constant is  . . . 180 ms. In contact with air, the conductivity increases up to s = 10-4 S/m owing to dissolution of CO2, leading to . . . 7.3 ms Therefore energy can be stored only for a rather short time in water-insulated systems, determined by the shorter of the two time constants for breakdown . . . and self-discharge." (Pulsed Power Systems Principles and Applications, Hansjoachim Bluhm (2006) p. 38,40;  Note: conductivity is given in Siemens per meter)

I found your question and following link when I googled dielectric constants of water at frequencies lower than 1GHz.

https://books.google.co.jp/books?id=pYPRBQAAQBAJ&pg=SA6-PA14&lpg=SA6-PA14&dq=water+dielectric+constant+10Mhz-2GHz&source=bl&ots=LTijl8zU3t&sig=ACfU3U0Gc0a1M6SkCk6rtuVKuHstzc6ONg&hl=zh-CN&sa=X&ved=2ahUKEwi45Jjfx6joAhWXHXAKHdYKBGE4ChDoATAAegQICBAB

It looks dielectric constant is almost frequency-independent but temperature-dependent below 1GHz. 80, 78 and 70 are good for 0, 25 and 50 degree Celcius.