Actually, if you look at the percentage chart, there are only three other recessions that had worse rates for the time period: 1948, 1953, and 1958.

Nobody really knows how many jobs are being created or lost at any given time. Those data are a combination of surveys and the silly requirement that businesses of a certain size report layoffs and plant closings. A sophisticated look at unemployment requires consideration of employment, unemployment, duration of job search, possible regulatory changes that affect the numbers and skill level of the unemployed over time (for example, states that have an inflation adjusted minimum wage have really done a number on their low skilled people), and the benefit packages that are available to support the unemployed as it is quite clear that higher unemployment benefits makes people more likely to extend their employment search.

I applaud the comments showing concern with the quality of the data being graphed. One of the biggest problems in the public debate is that beautiful graphics often shill for thoroughly uninformative or ill-considered data.