Combine different p.d.f. of the same random variable

Question

I have a random variable $X$, which I know the historical distribution (for 10 years I recorded this variable and I'm quite sure it follows this distribution) $f_h(X)$ that could be a Gaussian, Weibull or exponential. Then i have another distribution $f_f(X)$ that indicate a future forecast for this variable. I'm quite bad in statistic but i was thinking to:

• Use a convex combination where $f(X)=0.5*f_h(X)+0.5*f_f(X)$ with maybe different weight, in the picture is the "combined 1".
• Make a Joint probability distribution $f(h,f) = f_h(X) \cdot f_f(X)$ assuming that they are independent variables, in the pictures is the "combined 2" (here i had to divide the p.d.f. by the area under the curve to have a CDF=1)

Example pictures

thanks!