Abstract—Given the explosive growth of data that needs to be transmitted and stored, there is a necessity to focus on developing better transmission and storage technologies. The main goal of this paper is to develop the best compression methods for climate data. By using data compression, significant reduction in the bits to encode the climate data can be accomplished without loss of any important information. In this paper, humidity and precipitation data are considered for compression. The methodology is based on differential encoding wherein the prediction of the sample being encoded is obtained according to the output of an ANN (multilayer perceptron) whose inputs are time, month, temperature, pressure, incoming shortwave radiation, incoming long wave radiation, outgoing shortwave radiation, outgoing long wave radiation and solar radiation data for humidity and time, month, temperature, humidity, incoming short wave radiation, outgoing shortwave radiation, incoming long wave radiation, outgoing long wave radiation, and solar radiation data for precipitation. The ANN model uses 3 layers and 27 neurons for prediction of humidity data and 2 layers and 20 neurons for precipitation data. The highest compression ratio for precipitation data is 8.96 which is observed for the month of October and for humidity data the highest compression ratio is 6.92 and it is observed in the month of June.
Index Terms—Compression, compression ratio, artificial neural networks, climate data.
The authors are with the Department of Electrical and Computer Engineering, UNLV, Las Vegas, USA (e-mail: bharath.mummadisetty@gmail.com, astha.puri029@gmail.com, Shahram.Latifi@unlv.edu).
[PDF]
Cite:Bharath Chandra Mummadisetty, Astha Puri, and Shahram Latifi, "Lossless Compression of Humidity and Precipitation Data," International Journal of Computer Theory and Engineering vol. 9, no. 1, pp. 24-27, 2017.
