Context : Long-term sunspot observations are key to understand and predict the solar activities and its effects on the space weather .
Consistent observations which are crucial for long-term variations studies , are generally not available due to upgradation/modifications of observatories over the course of time .
We present the data for a period of 90 years acquired from persistent observation at the Kodaikanal observatory in India .

Aims : We aim to build a uniform sunspot area time series along with their positions , for the period of 90 years ( 1921-2011 ) , as obtained from the newly digitized and calibrated white-light images from the Kodaikanal observatory .
Our aim is to compare this new time series with known sources and confirm some of earlier reported results with additional new aspects .

Methods : We use an advanced semi-automated algorithm to detect the sunspots form each calibrated white-light image .
Area , longitude and latitude of each of the detected sunspots are derived .
Implementation of a semi-automated method is very necessary in such studies as it minimizes the human bias in the detection procedure .

Results : Daily , monthly and yearly sunspot area variations obtained from the Kodaikanal , compared well with the Greenwich sunspot area data .
We find an exponentially decaying distribution for the individual sunspot area for each of the solar cycles .
Analyzing the histograms of the latitudinal distribution of the detected sunspots , we find Gaussian distributions , in both the hemispheres , with the centers at \sim 15 ^ { \circ } latitude .
The height of the Gaussian distributions are different for the two hemispheres for a particular cycle .
Using our data , we show clear presence of Waldmeier effect which correlates the rise time with the cycle amplitude .
Using the wavelet analysis , we explored different periodicities of different time scales present in the sunspot area times series .

Conclusions :