The real plasma always interacts with the environment. In it there are directed streams of energy which is transmitted either by collisions of particles with each other or by radiation and absorption. Owing to it all the parametres of the real plasma are the functions of coordinates. The plasma can be not stationary. In the stationary plasma the power entering it is equal to the losses. What deviations from the thermodynamic equilibrium are in the real plasma - it depends at first on the correlation between the number of collisions per time unit and energy losses.

There are often states of the local thermodynamic equilibrium (LTE). So the state of the plasma is called by which all functions of distributions are balanced, exept one concerning the radiation. It means that as in the case of the full thermodynamic equilibrium the correlations (1) - (7) of the part "Radiation by the Thermodynamic Equilibrium" with the united parameter T are applied. But there is no equilibrium of the optic processes and Plank formula can't be used.

Тhe term "local equilibrium" came into being because it had to be brought into use the term "local homogeneity" of plasma for the description by this model of real inhomogeneous plasma. The plasma is devided into elementary homogeneous volumes (DV) and within each of them the parametres are statistically averaged. For this each particle inside the volumes each particle must have a lot of collisions. If the average free run of the particle between 2 collisions is designated as l_ef and considered parameter of plasma state as w then the condition of the local homogeneity will be expressed as:

The condition of quasistationarity of plasma is put down analogically. Within the time t_ef, which is equal to the average time of flying the particle between the collisions changing its impulse greatly, changing of the considered parameter must be little:

The correlations (1) - (6) of the part "Radiation by the Thermodynamic Equilibrium" with the same value T are applied to the plasma in which the condition (8) only for collision processes is fulfilled T. The value Т now is not the temperature in only thermodynamic meaning but it is a parameter close to the temperature value, which existed in the identical closed ensemble. LTE is typical for the most stationary plasmas received in the laboratory conditions.In the conditions of LTE of plasma the detailed equilibrium re. Optical transitions is upset, that's why it is expedient to consider radiation and absorption separately. The plasma in which the radiation of the given wave length practically isn't absorbed is optically thin for this radiation. The intensity of radiation J_ki of optically thin plasma, which is in the state LTE within a spectral line with the frequency n_ki is the following:

LTE plasma described by the united parameter Тcan exist in the limited area of pressure. The low limit of permissible pressures is set by the geometrical sizes of volume, where plasma is created, and the sizes must be much more than the value l_ef. The model of LTE of plasma is limited (from the side of high pressures) by pressures of some tens of atmospheres when the value l_ef becomes approximately equal to the average distance between the particles and the inequality (8) looses its validity. And in this case the plasma can't be considered as the ideal gas.

In the thermically not balanced plasma atoms, ions and electrons have got different kinetic energies, that's why they have got their temperatures, which can differ from each other very much. Different temperature names can be in the theory. Under the gas temperature T_г, or temperature of heavy particles is understood the characteristic temperature determined from the Maxvell's function of distribution (2): T_g = Mv²/3k. As the distribution of electrons on speeds can differ from Maxvell's much, especially in the area of high energies, the electrons can't be registed (given) any temperature. However by not very strong fields and not very little pressures the main part of electrons can be described by "balanced" distribution (2) with the temperature of electrons T_e = m_ev²/3k. The temperature of electrons T_e can be much more than the temperature of gas T_g.

Other typical temperatures can also be entered. The temperature of occupying T_o or the temperature of exitation of the given level T_e are calculated from the distribution law of Bolzman (4).

The occupyings of the group of levels on the regard for the occupyings of the main state can not answer the law (4), but the attitude of occupyings for any pair of this group of levels can answer the Bolzman law with the same temperature

In this case they say about the temperature of distribution T_d. The not balanced concentration of electrons or ions of the diven sort can be formally described by the equation of Sah (6) with the typical ionization temperature T_i.

For the plasma which is in the thermic equilibrium all the mentioned temperatures fit. This case is typical for the discharges by high pressure and big current. By the pressure ca. 1000 Pa and moderate currents (some amperes) the equilibrium is only between some states. So, exitation of atoms and ions occures mainly by the collisions of these particles with electrons of plasma. That's why the temperature of exitation of atoms and ions is practically equal to the temperature of electrons.

For example, in the laboratory conditions by the burning of arc discharge between metal or coal electrodes, the energy enters the plasma as Joul warmth by passing the electric current. Electrons get its main part, the electrons by collisions give the part of energy to the heavy particles - atoms. That's why in practice the equality T_g and T_е is followed not strongly enough. The method of relative intensities allowes to define the temperature of excitation and consequently the electronic temperature which is taken for the temperature of plasma (T_е= T_г).

Оne of the problems of plasma physics is to study the state of plasma by measuring its parameters: temperature, concentration of charged and neutral particles, distribution of different particles on excited states and also finding out space distribution of these parameters. Methods of investigation are united in the general notion of diagnostics of plasma.

Spectroscopic diagnostics of plasma is a research of plasma parametres on emited or absorbed by it radiation has important advantages. Main of them are - a lack of disturbances of the researched plasma and also the remote kind of measurements. The information contained in the absorbed or emited spectrum is very great.