What happens when you compare the Atlantic Multidecadal Index, Taymyr Peninsular tree ring reconstruction, the Central England temperature record, Volcanic Explosion Indexs and a few Siberian instrumental records? You crash the heck out of Opensource spreadsheets! You do get some interesting stuff though.
One of the biggest problems was that the AMO and the Taymyr had a pretty good correlation but a few things that looked goofy. The CET also match in a lot of places, but missed the boat in quite a few. So I averaged the AMO and the CET and compared to the Volcanic data available from the Global Vulcanism Program. The older major eruptions had a lot of estimates and questions on the dates. I just used the estimated date and best guess for the magnitude of the eruptions and used a 17 year moving average. Pretty crude, but it gives a rough idea of the impact of the Vulcanism on the AMO, mainly for the equatorial eruptions and the CET (March-April-May)/Taymyr for the northern hemisphere smaller eruptions.
The Average (1) curve is the average of the anomalies for three Siberian surface stations, Ostrov Dikson near Taymyr, Turuhansk, a little south of Taymyr and Pinofilovo which is in the major farming region of Siberia. The Pinofilovo station is plotted by itself to highlight how it tends to break the trends starting in around 1900. Far from conclusive that agricultural expansion is a major cause of the Siberian warming, but getting much closer to being somewhat convincing.
Russian records are pitiful, but there was a general expansion into the Siberian region at around 1900 and another push starting in the late 1950s. Russian production per hectare was about 30% to 50% less than in the US even though they exported more than the US in the early 1900s. One report for 2011 stated that Russian was planting 500,000 hectares of winter wheat. That is only 5,000 kilometers squared, but that should be about half of the acreage set aside for winter wheat allowing for crop rotation. There is also considerable barley, sugar beets and rye, for pasture and rotation, with about 8% to 10% of the total land area of Russia devoted to arable land. This is a smaller area than in the US but near the area in Canada used for grain crops. Canada also has a higher temperature trend in its interior farm belt than the global average.
The major albedo impact of large scale grain farming is two to three months per year, planting and harvest, where the land can absorb around 20Wm-2 more than it would in a virgin state. Since Russian records are so bad, I will have to use Canadian lands near the same latitude and attempt to interpolate the impact to Russia.
More: From Wikipedia, Total land under cultivation, 17,298,900km^2 or 11.61% of the land surface area. Russia, 1,192,300km^2, Canada, 474,681 km^2, United States, 1,669,302 km-2 and China, 1,504,350 km-2 totaling 4,840,633 km^2 which is 3.25% of the land surface area or 0.95% of the total surface of the globe. This doesn't include the second largest agricultural country India as I am mainly interested in impact above latitude 45 degrees North. This should include Kazakhstan and the Ukraine, with nearly 600,000 km^2, but I intend for their area to offset the southern portion of the US.
On a yearly average, the northern latitude albedo change due to agriculture would have to produce slightly over 1 Wm-2 to offset 50% of the estimated warming due to increased atmospheric CO2. Since the major impact of agriculture is only during the growing season and peaking in only about 2 months of that period, the impact during that time would need to be approximately 6 Wm-2.