Adsense earnings vary wildly between websites, niches, geolocation, even between days on the same site. It is very difficult to gauge performance by comparing two websites, unless those websites are very similar. Any comparison would require a lot of data over a long period of time (eg: month or year)/

is this too low?.

No. But it is really difficult to say and as CTR varies based on niche and other factors.I would say in general it is in the normal range.

How to improve this?

It is very difficult to make changes that will positively impact Adsense revenue. The price of ads is determined by the market through supply and demand. While one might think that websites or web-pages make up the supply side of the equation, they don't. The supply is driven by impressions. Say there are 1000 people looking up information on widgets, the supply is those 1000 impressions. Those impressions are occurring on any large number of websites. But, you say "my website is about widgets and I know the widget market and there are only 3 widget websites". That may be true, but AdSense shows ads based on the user's interest as well as the context of the website. So a user that visits your website may started to see ads about widgets only after they have left your site and are now reading news on CNN.com or reading their favorite blog about under water basket weaving. The point here is that widget ads can appear anywhere, and the ads that appear on your widget website are most likely not widget ads (many still are but most aren't).

Making changes to your website can improve performance but those improvements tend to be incremental. The only thing that has a significant and direct impact on ad revenue is increasing traffic.