a) Stratified random sampling, or simply stratified sampling. Each group individually is known as a stratum. The plural is strata. The key here is that each stratum is sampled, though we don't pick everyone from every stratum. We randomly select from each unit to have them represent their unit. Think of it like house of representative members that go to congress. We have members from every state, but Be sure not to mix this up with cluster sampling. Cluster sampling is where we break the population into groups or clusters, then we randomly select a few clusters in which every individual from those clusters is part of the sample.
b) Simple random sampling (SRS). This is exactly what it sounds like. We're randomly generating numbers to help determine who gets selected. Think of it like a lottery. A computer is useful to make sure this process is quick, efficient and unbiased as possible. Though numbers in a box or a hat work just as well.
For each of the methods mentioned, they aren't biased since they have randomness built into their processes.
We presume the temperature is decreasing at midnight, so reaches a low at 6 a.m.. The amplitude of the variation is (87-63)/2 = 12 degrees. We want the period to be 24 hours, so the argument of the sine function will be 2π(t/24) = πt/12. Then we can write
