So far I've only added a discussion of the graph.
It shouldn't need mention, but this answer is way beyond the scope of what you could do – it took me a solid hour to create and paste in the following. Thinking about data, what it means, and how to present it is obviously really important. I want you to demonstrate sensitivity to this. A bonus question is not the place for an extended treatment such as that below!
Here is the graph with the (i) seasonally adjusted and (ii) unadjusted data. I didn't include the slider, but add a second version that focuses on a particular time period to see the variation. I also added (iii) data in Chinese currency (officially the Renminbi 人民币 but more commonly simple the Chinese yuan 元). Because the exchange rate between the US$ and the RMB hasn't varied much, the two are almost the same. If you're Chinese, however, then "almost" isn't close enough, because that easily represents a 元10 billion discrepancy! For the purposes of thinking about trade issues in the context of Econ 102, the trends and levels are qualitatively identical. That would not be true if we picked a country pair where the bilateral exchange rate was volatile. Anyway, the monthly data are very volatile and so in this case using seasonal data makes it much easier to eyeball.
However, because this is a long time trend that goes from fairly small to very large, it's not possible to eyeball growth rates. There are two ways of handling that. One (duh) is to calculate a year-over-year growth rate. As we can see, there are some strong periods and some weak ones, but overall growth has slowed a lot and between early 2015 and the end of 2016 was on average negative. We can also see (i) the impact of the Asian Financial Crisis that began in July 1997 (and coincided at the start with a recession in Japan). We also see the Great Recession in 2008-2009 and slow growth following the collapse of the dot.com bubble in 2000. The 1996 slowdown reflected domestic Chinese factors, including an appreciation of the RMB.
Another way of presenting data over a long time trend is a log scale. This helps us eyeball periods of rapid and slow increases, and (unlike with growth rates) gauge the cumulative impact, which is a substantial overall increase. If we look at the graph, we can see that (i) there wasn't much change in 1992-93, a big jump (1 year for one increment) in 1994, then slower growth (5 years for one increment), then 2.5 years (through early 2003), 1 year and then 2 years for the next 2 increments, then 4 years (through early 2005), and now after 12 years exports have yet to grow another increment. Now the dollar amounts are large,
since the total is near enough $200 billion dollars, but the change is small in percentage terms. Whether you want to look at the former or the latter is a function of why you're looking at the data.
Then there's the question of price increases and overall economic growth. One way is to look at exports to China versus China's GDP. One awkward aspect is that the data aren't both seasonally adjusted – FRED only provides a handful of GDP series for China. But while we see a rise and fall, the bottom line is that most of the long-run rise in US exports to China reflect the growth of the Chinese market. How could it be otherwise? Well, China could be importing more overall, so growing imports amplified by a growing economy. This graph by the way uses nominal GDP in RMB and nominal exports in RMB.
Next we can look at US exports to China relative to total US exports. In 1992 China was still very poor; not today! So we would expect exports to comprise a modest share at the start, and rise over time – which is what we observe. In 1992 only 2% of our exports went to China. Now it's 15%.
Finally, there's the politically important question of bilateral trade. As an economist, I'm fundamentally not interested in that. Instead I want to know whether aggregate US trade is balanced. But those inside the Beltway, and most of our news media, don't view things that way. Of course this graph is seasonally adjusted, but not with a log scale...you now know how to adjust.