How do we get from $(1)$ to $(2)$ ? $$\frac{\partial}{\partial x} \frac{f(x,y)}{G(y)}+\frac{\partial}{\partial y} \frac{f(x,y)}{G(y)}=0 \tag{1}$$ $$ \frac{f(x,y+z)}{G(y+z)}=\frac{h(x-z,y)}{G(y)}, \text{for } x \geq z\tag{2}$$
-
Why is equation $(1)$ written using total derivatives rather than partial derivatives? Is it really supposed to be that way, or did you just not discover how to show $\frac{\partial}{\partial t}$? – David K Sep 24 '20 at 12:19
-
No, it should be partial derivatives. – Cookie4free Sep 24 '20 at 12:21
-
OK, that makes much more sense. I put $\partial$ symbols in the formulas. You can edit it further if it is not correct. – David K Sep 24 '20 at 12:25
-
Thank you. No, now it is correct, but I still don't understand how I get the simplification. – Cookie4free Sep 24 '20 at 12:27
1 Answers
First of all, we have to recognize that we are treating $\frac{h(t,x)}{N(x)}$ as a single function of two variables. Let's write this explicitly:
$$ f(t,x) = \frac{h(t,x)}{N(x)}. $$
Now the fact that $$\frac{\partial}{\partial t} f(t,x)+\frac{\partial}{\partial x} f(t,x) = 0 \tag1$$ tells us that the directional derivative of $f$ in the direction $(1,1)$ is zero. That is, if you increase each of the two input variables of $f$ always at the exact same rate, the value of $f$ never changes. After doing some amount of increase to each variable in this way, you will have increasing each of them by the exact same amount without changing the value of $f$.
So for some values of $t,$ $x,$ and $z,$ we start with $f(t - z, x)$, which has some value, and we increase the first and second input variables by the same amount, $z$, arriving at $f(t, x+z)$, which produces the same value. For this to work, however, $f$ must be defined at each of the input pairs $(t-z,x)$ and $(t,x+z)$ and Equation$~(1)$ must be true at every point on the straight line segment between those two points (interpreted as Cartesian coordinates), that is, Equation$~(1)$ must be true whenever we replace $(t,x)$ with $(t - z + \eta, x + \eta)$ for $0 < \eta < z.$
The condition $t \geq z$ does not come directly from Equation $(1)$ but it may come from some limitation on the domain over which Equation $(1)$ is true.
For a more mechanical proof, for a given $t$, $x$, and $z$ you can define $u(\eta) = t - z + \eta,$ $v(\eta) = x + \eta$, and $$ g(\eta) = f(u(\eta), v(\eta)). $$ Then $$ \frac d{d\eta} g(\eta) = \frac{\partial f(u,v)}{\partial u} \frac{du}{d\eta} + \frac{\partial f(u,v)}{\partial v} \frac{dv}{d\eta} = \frac{\partial f(u,v)}{\partial u} + \frac{\partial f(u,v)}{\partial v} = 0.$$ Now simply integrate $$ g(z) - g(0) = \int_0^z \frac d{d\eta} g(\eta)\,d\eta = 0. $$
Hence $f(t,x+z) = g(z) = g(0) = f(t-z,x).$
- 98,388