About
The project aims at acquisition, manipulation, visualization and some rudimentary analysis of a set of data involving precipitation, streamflow, temperature and drought in California during the period of 2012 to 2019. This period is identified to be the recent prolonged drought in California. Earlier two significant droughts were reported from 1928 to 1934 and 1987 to 1992. But, the analysis is restricted for the recent drought due to data availability and relevance of time frame. Sufficient data related to streamflow has been obtained through USGS Surface Water portal, and precipitation and temperature data through NOAA for the prolonged period. The drought data has been obtained from the United States Drought Monitor (USDM) website. The data itself is an area percentage in floating value of the state under the influence of five different categories of drought severity. Exploratory graphical analysis of the data showed some good trends between the parameters and the drought area. Different machine learning algorithms have been evaluated against the actual drought data and depending upon the accuracy, a suitable algorithm for each intensity has been determined.
