Dear Zainal

Try to use XL Stat software. XL-Stat have trial version for 30 days. You can use it for your data.  The tests you mentioned are available there. 

Thanks 

I am not sure which homogeneity test you are trying to preform: means, variances, ...?

R does have all of these tests. I would not recommend XL-Stat for this.

tq sir A.K. Singh for your reply. for your information im trying to perform the means of rainfall data as due to my reading on those test i mention above, they test mean of annual rainfall. Or sir did you have any other suggestions on how i would test for the homogeneity in may daily rainfall data?

The common test of homogeneity of means (assuming there is just one FACTOR affecting the response) is 1-WAY ANOVA which assumes that (1) the error variance is same across all FACTOR levels, and (2) the error is normally distributed. When the variances are unequal, it is better to use bootstrap method to calculated P-values from ANOVA. Not knowing what your data looks like, this is all I can tell you. If I can see the data, I can give you better advice.

sir A.K. Singh,

here i attach my data sheets. my data is a daily rainfall data. so i need to test the homogeneity of the data that were collected from a station in which to know whether the data are doubt or not.

I am confused about which homogeneity test you are trying to preform

I'm currently working on this as prerequisite for my own data analysis too, in fact a Google search led me here.

If you are comfortable using R, do a Google search on RHtests_dlyPrcp, RHtestsV4, and RClimDex.

Another GUI based programme is AnClim/ProClimDB. It has a good graphical interface but you have to request the full version to work with daily data.

Hi Oluwaseye Agunbiade...

May I know.. if u are not hardly to share with me... may i know which one from the above u mention that u prefer with? 

I started with Anclim (which I preferred because it seems to have more functions) but I just found out that the free version does not work with daily data. I have requested for the full version and I'm still awaiting the reply.

Meanwhile I have gone back to trying out RHtests_dlyPrcp, I'm still working on it.

Dear Deana Zainal,

definitely, I would recommend you the RHtests which were mentioned by Oluwaseye Agunbiade. More specifically, I would draw your attention to the RHtests_dlyPrcp R script that you can be given on request directly from its creators (for contact details see the first link). In 2014, they replied practically immediately. If you are interested in a more detailed description of this script, read the original paper by Wang et al. (2010; attached paper). I have not used their GUI, I always run the computation directly from the command line. Among the outputs you can also find the homogenized series. Note furthermore, you can also check the homogeneity of daily discharge series. Recall, for instance, that if you have the catchment area, you can easily compute the runoff depth corresponding to your prescribed time periods such as days, months or years (i.e. an equivalent of precipitation).

Second, I would like to react to the Oluwaseye Agunbiade's reply on AnClim. I am not aware of the fact  that AnClim is able to work with daily data as such. Instead, I would expect that, when dealing with daily data, it first summarizes them to monthly data or it proceeds as done in the paper of Wijngaard et al. (2003; second link). However, I haven't worked with the professional version. And you, Oluwaseye? Did you get some response from P. Stepanek?

Third, recently, when working with AnClim, I have noticed that there may be some mistakes regarding tabulated critical values. I came across some of them when comparing homogeneity tests included in AnClim and those ncluded in the R package called 'climtrends'. If I remember well, it was the Buishand test (aka 'Cumulative Deviations' in AnClim). I really regret saying this, but I cannot recommend AnClim at this time even though it is a product of our institute.

http://etccdi.pacificclimate.org/software.shtml

http://journals.ametsoc.org/doi/abs/10.1175/2010JAMC2376.1

http://onlinelibrary.wiley.com/doi/10.1002/joc.906/abstract;jsessionid=A93719557FB419BB8B69D5971ACD3ABA.f03t02

Homogeneity in data sets can be checked using graphical as well as statistical approaches. The most common graphical approach is the "Double Mass Curve" technique, which indicates the change in slope of data as break point from which the data exhibits inconsistency.

Secondly, the statistical tests such as Buishand test, Standard Normal Homogeneity test (SNHT), Pettit test, Mann Whitney Pettit (MWP) test, etc. can be used. You can refer the book "Hydrological Time Series Analysis: Theory and Practice" by Machiwal and Jha for details of these tests.

You can download XLSTAT software which is an extension to MS Excel. It has all features inbuilt for time series analysis. It will give you wide range of statistical testing.

Download RHtests_dlyPrcp.r from weblink

http://etccdi.pacificclimate.org/authenticated_downloads/RHtest/RHtests_dlyPrcp.r

Load source R code RHtests_dlyPrcp.r and then execute thereafter for the following mentioned tests.

• Except VoR test, all other tests are available in "trend" package of r.

https://cran.r-project.org/web/packages/trend/trend.pdf

Oluwaseye A. Agunbiade Would you like to share RHtests software as mentioned at following address

http://etccdi.pacificclimate.org/software.shtml

Regards.

Naveed

you can use Anclim software to use all these tests except VoR test !