If $\int f=0$ and $f(x) \ge 0$ for all $x \in \mathbb{R}^d$, then $f=0$ a.e.
I let $E \subset \mathbb{R}^d$ be a finite measurable set.
I try to break this into two cases:
Case 1: If $f(x)=0$ on $E$, then we are done (as trivially $f=0$ a.e.).
Case 2: If $f(x) > 0$ on $E$, then $\lim_{n \to \infty} \sum_{k=1}^n a_k m(E_k)=0$. I think I can prove by contradiction here. Can I assume $f \not= 0$, which means $f > 0$ in this case? Then, do I need to show that $\int_E f > 0$, which would contradict the hypothesis of $\int_E f = 0$. Is this a good approach to this?
Here's another way to do it which doesn't explicitly make use of approximating $f$ by simple functions.
For $n \in \mathbb{N}$, let
$$E_n = \left\{x \in \mathbb{R}^d \mid f(x) \geq \frac{1}{n}\right\} = f^{-1}\left(\left[\frac{1}{n}, \infty\right)\right).$$
Note that $\displaystyle\bigcup_{n=1}^{\infty}E_n = f^{-1}((0, \infty))$. If $f$ is not zero almost everywhere, then $m(f^{-1}(0, \infty)) > 0$ so there is $n \in \mathbb{N}$ such that $m(E_n) > 0$. Now note that $f \geq \frac{1}{n}\chi_{E_n}$ so $$\int_{\mathbb{R}^d}f\, dm \geq \int_{\mathbb{R}^d}\frac{1}{n}\chi_{E_n}\, dm = \frac{1}{n}m(E_n) > 0.$$
Therefore, if $\displaystyle\int_{\mathbb{R}^d}f\, dm = 0$, $f = 0$ almost everywhere.