Why is the engine suggest Bd3?

Hello,

In the below it's white's (me) turn to move.

I played Bd2 to protect my knight, but the engine called it a blunder and recommended Bd3 as a best move. I don't get why. There is no threat on D3, and black could then play Bxc3+ which I would then have to counter with xc3 and end up pawned...

Am I missing anything?

I don't think the engine would call Bd2 a blunder. (Which is a 2 point negative swing)

But the bishop has better places to go.

You can see from the above that few master games see it. (8 games for 5 Bd2 compared to 1944 for Bd3)

They are not worried about black giving up his bishop for the knight.   Bd2 seems a passive spot for the bishop. White is trying to use the first move advantage. 

As far as the computer eval, you can read them to compare the thousands of lines it was looking at. 

But wouldn't pawning on the c column be a weakness?

Yes, doubled pawns would be a bit of a weakness, but having the two bishops is a bit of an advantage.  To have black surrender the two bishops without your having to spend a move with a3 to force it is a slight net benefit for white.  Also, Bd2 doesn't really improve things, as after bxb bxb white's bishop needs to move or its a big pawn.  Its a hefty price to pay to avoid doubled pawns.  

Thanks ArtNJ, that makes sense!

And the isolated pawn on a2 is not much of a weakness.

Also the doubled c-pawns are not a permanent weakness, white can play c4 later and I don't see a good way for black to stop that. 

For example Bd3 Bxc3 bxc3 Nf6 Nf3 0-0 0-0 Re8 c4. c4 may not even be best there, maybe play it a few moves later. The point is Bxc3 is a bad move. Black can only stop c4, if white allows black to.

"Why is the engine suggest Bd3?"

Since i never actually answered your question, I will now.  Because it places the bishop on an active diagonal.  Count how many squares the bishop controlled while at f1, and then compare that to how many squares it controls on d3.

Of course 5.Bd3 is the most sensible move, but there is also Larsen's old move 5.Qf3!? which makes sense.

not really. the c-file is not open so black would not have an easy time exploiting it anyway. plus black would be giving up the bishop for the knight - A bad trade, and giving you a half-open b-file which would be a nice address for one or both of your rooks.

What’s wrong?

since there is a d-pawn, the doubled c-pawns are actually an advantage as they can give you better center control and the rook on a1 will have the b-file to use.

Yes. Doubled Pawns are not always a drawback, and Black is giving up his good Bishop in order to create the doubled Pawns.

The main drawback of doubled Pawns is that they tend to reduce the collective mobility of the whole Pawn constellation that they form part of. Doubled Pawns are usually weak if they can be turned into an immobile target. That's not the case here, since it seems trivially easy to undouble the Pawns with a later c3-c4.

Also, the doubling of the Pawns, transferring White's b2-Pawn to c3, gives White an open b-file and more central influence. The loss of Black's dark-squared Bishop also leaves his dark squares on the K-side and in the center rather weak.

Black's Bb4 move isn't a blunder perhaps, but it's not particularly promising either. White's Bd2 looks like a waste of time, countering an imaginary threat.