I know you've deceived me, now here's a surprise I know that you have 'cause there's magic in my eyes I can see for miles and miles and miles and miles and miles Oh yeah The Who, I Can See For Miles

Assignments: Read Chapter 24, Sections 4 and 5

In Class: ------------------ flux is what we can measure directly for celestial objects - can't measure luminosity (need to know how far away you are) We would like to know the luminosities of stars, so how can we do this? 1. calculate the DISTANCE using the method of parallax 2. calculate FLUX by measuring it directly 3. INFER LUMINOSITY from FLUX and DISTANCE using relation 4. Results: a) stars are really LUMINOUS and DISTANT -- they are distant Suns -- not little nearby comets b) the luminosities of stars vary over a HUGE range -- less than 1/10th the Sun's LUMINOSITY -- more than 10,000 times the Sun We've answered one question (Are stars distant Suns?) but in the process generated another, even more puzzling one (Why don't stars all have the same LUMINOSITY?) -- they must not all be the same as the Sun -- so how do they differ? -- how do the differences translate into different LUMINOSITIES ------------------------------------------ Luminosity is not the only property of stars we can determine Color varies as well -- though all stars are pretty much white-ish, there are readily discerable color variations. -- Aldebaran, Betelgeuse are very red stars -- Rigel, Sirius are very blue stars -- the Sun is a fairly yellow star "Color" means that one color light is brighter than another color for a given object -- blue stars emit more blue light and less red light -- red stars emit more red light and less blue light We can measure color by measuring the brightness of stars in different colors. -- use filters -- block all light except light of a given color -- compare "red" and "blue" brightness -- maybe use ratio "red"/"blue" to assess color -- red stars -- brighter in red than in blue -- big value for ratio -- blue stars -- fainter in red than in blue -- small value for ratio We can set up a "COLOR SEQUENCE" -- order the stars based on their "redness" -- color is an intrinsic property of stars -- ratio of two FLUXES -- therefore ratio of two LUMINOSITIES -- not dependent on DISTANCE from star -- e.g., a star looks the same color not matter where you're located Now we have LUMINOSITY and COLOR -- intrinsic properties for every star we measure -- give astronomers two quantities, and they'll make a graph -- LUMINOSITY vs. COLOR -- every star can be put on a graph -- every point on the graph denotes a specific (COLOR,LUMINOSITY) pair -- red stars to the right -- bue stars to the left -- high LUMINOSITY stars to the top -- low LUMINOSITY stars to the bottom -- relationship between LUMINOSITY and COLOR?? -- could be unrelated -- then there would be no trend evident -- red stars both of high and low L -- blue stars of both high and low L -- high L stars both red and blue -- low L stars both red and blue -- scatter diagram -- if these quantities are "correlated" -- should see a trend in the graph -- points NOT evenly distributed all over -- clustering -- red stars are usually of low L -- blue stars usually of high L -- if you're told one quantity, you can pretty much guess the other -----ASIDE on CORRELATIONS---------- - is there a relationship between two quantities just because there's a correlation? - e.g., # of policeman vs. crime rate - if so, then maybe our $$ spent on more policemen are justified - we can predict that the crime rate will go down if we have more policemen - if no correlation, then it doesn't matter how many cops are on the street - maybe we should spend our $$ elsewhere - sometimes a correlation doesn't mean a relationship - e.g., # of priests vs. # of murders - they're probably not causally related (I hope) but both might result from an increasing population - The existence of a correlation does not necessarily imply to two parameters are CAUSALLY related, but the absence of a correlation does indicate that the two parameters are unrelated. Correlations are a necessary but not sufficient indication of a relationship -----------END ASIDE----------------- -- when you plot the "real" data for "real" stars -- there is a good correlation -- stars clustered along a line in the graph -- relationship between LUMINOSITY and COLOR -- red stars typically of low L -- blue stars typically of high L