Greetings, 

please check this paper: 

https://pdfs.semanticscholar.org/7d4d/069663de47e11fd769bb267b375bce6d5f31.pdf

Hi, I can add that the density of meteorological stations depends strongly on the topography of the area.

For example, in the following article you can read about the spatial variability of temperature:

http://www.ccsenet.org/journal/index.php/jas/article/viewFile/23956/15205

Dear Dr. Nawbahar ;

First of all make sure that around your numerous station there is no obstacles which could affect the quality of the measures (like topography, high building that affect the free circulation of wind or the arrival of correct solar radiation...).

Generally, station must be placed away from all factors which can effect measures and the highest regarding to the height of each buildings or trees in agricultural areas.

Nawbahar,

Please be specific about the objective and the variables you are interested in. The density of the network depends on the objective of your study and variables you are interested in. If you are just looking for only tempetature and winds. sparse resolution would be sufficient, unlike, rainfall, which is a quasi random, and needs high resolution data. As Sonia mentioned topography matters. If the surface is heterogeneous, variance of the boundary layer variables are high, and possibly with shorter correlation length scales.

One of the important requirement of right location is to make sure that the flow is stationary, which means, terrain should be homogeneous, and far away from the obstacles. Rule of thumb would be have a ratio of fetch (horizontal distance from the tower towards the direction of wind) to the height of the instrument location should be at least 10.

You may want to refer to one of my papers and references their in which discuss some of these issues.

http://link.springer.com/article/10.1007/s12040-008-0076-0

Hope it helps...

Good luck !

To complement Arunchandra, there can be no errors in locations' selection of the stations, but they can be errors in your conclusions based on your data analysis! Stations will measure the parameters they measure as long as instruments work according to specs, what this measurements mean to you is another thing!

Nawbahar, please try to specify your problem, otherwise the best answer is - all density and all location are just fine :)  

Thank you 

Dear Nawbahar Mustafa,

The optimal density of meteorological stations depends on the spatial-time scale atmospheric circulations that we want to observe. For large scale, such as the general circulation of the atmosphere, the lower density of meteorological stations is needed, while for less scales, such as synoptic circulations (atmospheric fronts and extratropical cyclones), requires a higher density of meteorological stations. The highest density of stations is necessary for at least the circulations, such as, for example, tornadoes. So, the application of the principle of scale analysis gives the best results in the optimization of the density of meteorological stations. Theoretical and practical bases of optimization can be found in the famous book Gandin (Gandin, L. S., 1963: Objective Analysis of Meteorological Fields. GIMIZ, Leningrad. (Translation by Israel Program for Scientific Translations, Jerusaleum, 1965)).

Best,

Milivoj B.Gavrilov

Thank you.

Milivoj is correct providing all used stations are representative of their locations or at least comparable. This is generally the case with the networks of National Meteorological Services that should be in accordance with WMO standards (e.g. WMO Basic Docs,  subsection 2.4.1.1). But since Nawbahar still didn't define her particular problem, this assumption may prove to be wrong, e.g. if using the data from a research network or amateur meteo networks, or some other third party source. 

Just saying... 

http://library.wmo.int/pmb_ged/wmo_49-v1-2015_en.pdf

Thank you

Dear Nawbahar, 

The number of monitoring stations and their location or distance from each other depends on the aims of the network and what you want to achieve from the data. Do you want to monitor micro climate changes or regional and global. Do you want to know micro level factors affecting the meteorology or regional level? Do you want to monitor for the purpose of health effect? All these questions and many more will deifne the density and location of your monitoring stations. A very important question is, how much funds are available for setting the network? 

Any good luck for your monitoring and research program. 

Thanks