[origin_path]

Because that's where I happened to find the graphs? Go right ahead and find the papers if you like. The research in question is simply plotting old model projections on a graph along with satellite temperatures.

"Wouldn't it make more sense to get a variety of sensor data?"

By all means, try it. You'll find that satellites and weather balloons disagree with surface level temperature history. The reason is, climatologists keep editing surface temperature databases to create new 'versions' that create warming trends where none were previously visible. This is not scientific, of course. You're meant to fit your theory to the data, not data to the theory. The satellite data presented there is much less tampered with.

[hotpotamus]

How do I know the satellite data wasn't tampered with?

[origin_path]

Good question! In fact some satellite datasets are tampered with. There's a good example documented here:

https://www.youtube.com/watch?v=bOHrYY3yAGE

The RSS dataset showed no warming for nearly twenty years - a problem, it's not meant to show that. Once the data was released, people with experience with how climatology operates (like Dr Spencer) predicted the data would soon be revised to show warming, somehow. That's exactly what happened. The original data used confidence intervals that widened over time, to reflect degrading instruments and orbits. In the new data the CIs were gone and the new trend line was simply the uppermost value allowed within the CI for each point. Because the CIs were widening, this trick created an appearance of warming where previously there hadn't been any.

Generally, the only way to be sure something isn't being messed with is to review the methodologies. The tampering is public, it's not a secret and they don't hide the fact that they do it, they just don't mention it to the general public and rely on the media to not report what they're doing. It works: awareness of this problem is incredibly low. Still, you can compare the different 'versions' they release over time to see the enormous magnitude of the changes, or you can read the occasional reports that surface in outlets like Nature when really massive revisions occur, and then you can remember that scientists aren't meant to do that. They're meant to add error bars if they have doubts about their data and propagate that uncertainty through to their final graphs and predictions. Climatologists don't do this. Instead they argue that they know temperatures should be going up, so if it isn't, there must be something wrong with the data. Nothing a quick model can't fix.

https://www.nature.com/articles/nature.2015.17700

"An apparent pause in global warming might have been a temporary mirage, according to recent analysis. Global average temperatures have continued to rise throughout the first part of the twenty-first century, researchers report on 5 June in Science. That finding, which contradicts the 2013 report of the Intergovernmental Panel on Climate Change (IPCC), is based on an update of the global temperature records maintained by the US National Oceanic and Atmospheric Administration (NOAA). The previous version of the NOAA data set had showed less warming during the first decade of the millennium."

[hotpotamus]

Yes, but how do I know the original data wasn't faked to begin with?

[origin_path]

You don't but this is a classic problem to do with proving guilt vs innocence, isn't it. You have to look for explicit evidence of tampering, instead of trying to prove a negative. That is, start from the assumption that the various actors began in a well meaning place, and then react to evidence of wrongdoing, rather than assume wrongdoing from the start and then try to prove innocence. Otherwise you can never bottom out.