ACF and PACF Plot interpretation for ARIMA

[Fitch92]

Hello,

I just plotted my ACF and PACF Plot after differentiating the time series.

I as well used the auto.arima() function to get the final model. I therefore received:
for AIC:

ARIMA(0,1,2)(1,0,0)[52]

for BIC:

ARIMA(1,1,1)

I actually have no idea how one should read the plots to get these results. For me its a little contrary.
Can you help me?

Thank you in advance.

[englianhu]

??forecast::auto.arimawill auto use the best c(p, d, q)

Here is example binary.com 面试试题 I - GARCH 模型中的 ARIMA(p,d,q) 参数最优化

[Fitch92]

I know, but I actually don`t get how forecast::auto.arima gets these values just from looking at the ACF and PACF plots