Arguments
Air pollution and climate change could mean 50% more people going hungry by 2050, new study finds
Posted on 8 August 2014 by Guest Author
This is a re-post from Roz Pidcock at Carbon Brief
The combination of rising temperatures and air pollution could substantially damage crop growth in the next 40 years, according to a new paper. And if emissions stay as high as they are now, the number of people who don't get enough food could grow by half by the middle of the century.
Burning question
Feeding the world's rapidly growing population is a serious concern.
Research shows rising temperatures are likely to lead to lower crop yields. Other work suggests air pollution might reduce the amount of food produced worldwide. But nobody has considered both effects together, say the paper's authors.
The two effects are closely related as warmer temperatures increase the production of ozone in the atmosphere. Here's lead author Professor Amos Tai from the Chinese University of Honk Kong to explain.
The new study looks at global yields of the four principle food crops - wheat, rice, corn and soybean - and how they're expected to change by 2050 under different levels of future emissions.
Together, these provide nearly 60 per cent of all the calories consumed by humans worldwide.
Global losses
The maps below show some of the results.
The top panel is an optimistic scenario in which greenhouse gases stabilise at 630 parts per million (ppm) by 2100. For reference, we're at about 400 ppm now.
The team compared this with what might happen if greenhouse gases continue to rise as rapidly as they are now. That's the bottom panel.
How global crop yields are expected to change between 2000 and 2050 for RCP4.5 (top) and RCP8.5 (bottom). Total annual crop losses for each scenario compared to today are in the purple boxes below. Source: Tai et al. ( 2014).
Global food production falls by 2050 in both scenarios, as shown in the purple boxes below each map. If emissions stay high, total global yield falls by about 15 percent compared to today. But if emissions are kept low, the drop is reduced to just over nine percent.
The coloured shading shows how the total yield for all four crops varies region to region. Red is a higher yield than now, blue is lower.
Food security
Does this mean more people won't get enough to eat?
If emissions continue as they are, the drop in crop yields will means many more people globally will be undernourished, as the graph below shows.
The blue shading is the fraction of people in developing countries who consume less than the minimum amount of energy required per day. At the moment it's about 18 per cent.
The red and blue shading together is the fraction of undernourished people by 2050 under the high emissions scenario. It's about 27 per cent - about 50 per cent more than now.
The projections include expectations of a 40 per cent rise in global population by 2050. But they don't take into account measures to increase resilience, such as changing farming practices.
Are we heading for global famine?
A headline in yesterday's Daily Mail suggested the new research meant "climate change and air pollution will lead to famine by 2050". But Tai tells us that's a "little bit of a stretch".
Famine is more complex than just food shortage, he explains:
"Famine is an extreme form of undernutrition with significant fatalities caused by not only crop failure but also a variety of socio-economic factors, which our study did not examine. The problem of undernourishment, in a more general sense, is usually more subtle but widespread and persistent".
Different crops, different pressures
The scientists found not all crops are affected in the same way. For example, corn is more susceptible to heat, while ozone exposure is more of a problem for wheat.
How badly different countries are affected will depend on how bad air pollution is already and which crops they grow, the paper explains. For policymakers trying to decide how to deal with the problem, different countries may need to try different approaches. Tai tells us:
"Depending on the crops of interest and where you are, you may want to focus more on air pollution reduction or climate adaptation more."
For example, tougher air quality regulations in the United States could lead to a sharp decline in ozone production, lessening the impact on crops.
"Talk to each other"
The authors are calling for greater collaboration between agricultural planners and air quality managers to set goals for food security and public health. Tai tells us:
"For policymakers and people in the field, including farmers, air pollution managers … they do have to talk to each other … We really call for greater collaboration between stakeholders of different interests and fields … they cannot just be concerned about their own fields because their objective is ultimately affected by decisions made in other sectors."
Scientists know cleaning up air quality will have a significant impact on public health. This new research suggests cutting domestic air pollution as well as curbing warming could have big benefits for food security in some parts of the world too.
Let's also not forget that if it came down to it we can stop using corn to make fuel.
Michael. ... when someone is trying to justify why nature is behaving a certain way as is the case with global warming it can be very blinding to get caught up in the minutia of individual studies especially when studies conflict each other. I have read hundreds of climate studies over the years. I have a degree in environmental biology so I know a thing or two about succession. You asked me why I believed that the climate models were inaccurate and I provided 2 different studies for you. I don't think that the models are useless. ... I agree that they can help us continue to refine the complex mechanisms that dictate climate. However no matter how much you insult or belittle me I still am skeptical about the reliability of using the models to predict the climate. But if it makes you feel smarter to belittle me.... go ahead.
Donny wrote: "...I still am skeptical about the reliability of using the models to predict the climate."
So don't.
Ignore the models. Look at analyses of past climate changes. Or the measured impacts of the current ongoing climate change. Or the pure physics calculations of the effects of increasing greenhouse gases and directly linked feedbacks.
These all point to the same general conclusions as the models. The primary benefit of the models is in allowing us to estimate different future scenarios based on our actions in the upcoming decades. Some are also starting to provide reasonable estimates of regional, rather than global, impacts. However, if we tossed the models out entirely all available evidence would still be indicating that the rapid CO2 increase we have introduced is causing the rapid warming we are observing.
You don't need models to predict the climate. Arrhenius successfully predicted many of the climate changes which have now taken place all the way back in the 1890s... decades before modern computers, let alone climate models, even existed.
'Models aren't perfect' is true, but generally a poor excuse for ignoring them. However, as an excuse for ignoring the over-whelming evidence that has nothing to do with models it is just plain illogical.
Donny @52.
Surely the reason why you "still (are) skeptical about the reliability of using the models to predict the climate" is mainly because you still consider the "case with global warming" to be "nature ... behaving a certain way." That is, you are simply in denial over AGW, and that is despite your studies in Environmental Biology.
Concerning the topic of discussion you presently find yourself engaged in here - the impact of future AGW on agricultural output - from your studies you should know that primary succession requires a lot of time for soil formation. Thus where soils are presently inadequate for agriculture, any profound allogenic change due to AGW that acts over just a few decades will not alter the agricultural usefulness of soil, unless perhaps we take up farming lichen & moss (yum yum!!).
But, hey, @50 you seem to be saying that a landscape free of glaciation for around 12,000 years will result in soils with which "we shouldn't have much of a problem." Now, I say "seems to be saying" as I am but positing the idea that this isn't another instance of Donny "responding flippantly." It may well be otherwise as even when glacier-free for many millennia, the lands north of Canada's prairies are not renowned for deep soil cover.
From a few months ago at http://thinkprogress.org/climate/2014/03/07/3370481/california-drought/
"… scientists a decade ago not only predicted the loss of Arctic ice would dry out California, they also precisely predicted the specific, unprecedented change in the jet stream that has in fact caused the unprecedented nature of the California drought. Study co-author, Prof. Lisa Sloan, told me last week that, 'I think the actual situation in the next few decades could be even more dire that our study suggested.'"
(Thanks to hank at RC for this quote and link.)
(I have concluded that 'donny' is a sockpuppet of one of the mods trying to yank our collective chains--good one, guys!)
[PS] fixed link ... and we wish....
Bob Loblaw @47, your first link is indeed the map I was indicating. I have since found a more suitable map (the third map from this site to which you have previously linked), as it shows the actual soil qualities. Of particular interest in your dispute with Donny is the large swathes of soil in the east of Canada which have "No Capability for arable culture or permanent pasture" (pink) despite being at the same latitude as the Canadian prairies, and hence are not restricted by climate factors.
Equally interesting are the areas of arable land (some of the highest quality) along the Peace river (north west Alberta) and the single mapped section in north east British Columbia showing a complex interlacing of arable soils (brown) adjacent to a large section of peats (black) which are unsuitable for agriculture. The high latitude of these sections relative to the northern limit of the Canadian prairies shows that northern limit to have been set by soil type rather than climate.
Those examples, however, also show where, and how we can expect to find suitable soils in Northern Canada with increased global warming. Both represent deposits of river silt, a process that produces suitable soil for agriculture almost regardless of climate (ie, soil in which agriculture is only restricted by climate). There presence indicates that such suitable soils are also likely to be found in the river valleys of the MacKenzie and other smaller northern rivers descending from the Rockies. Comparison with the soil map of a Canada (first map, previously linked site) does show the relevant reaches of the McKenzie have luvisolic soils, and thus they are likely to provide small regions of class three or four soils (as with BC reaches of the Peace, which has a similar soil classification). That is, the land will be potentially arable, but will require extensive conditioning to control soil pH (if I understand the description correctly).
As an aside, a more detailed soil map of Canada, and indeed of Siberia as well, can be found in the "Soil Atlas of the Northern Polar Region" (280 MB PDF; website). Unfortunately the basis of classification in that atlas does not lead directly to an ability to classify as to whether or not the soil is arable, at least for inexpert commentors like me.
With regard to Mills (1994), closer reading shows that he finds 16 million Hectares of arable land that would be opened up from a doubling of CO2, but that the majority (15.9 million Hectares) of that will be in Alaska, while only 0.2 million hectares will be within the study area within Canada itself (page 122). The study area within Canada is quite restricted (fig 1), so that some more might be expected in Canada, but arguably not a lot as the most suitable sites on geographical grounds are also the most likely to have been surveyed. This difference between Alaska and Canada is consistent with the hypothesis above that river valleys will produce arable land even in climates currently to cold to sustain cropping, along with the size of the Yukon river valley.
I am not sure if this additional information reconciles you to the Mills (1994) results, but regardless I will accept the result of a peer reviewed study over the anecdotal evidence of a non-expert 100% of the time. If you think Mills is wrong, find a more recent peer reviewed study showing that he is wrong.
Tom, I think the wider range of soils in Alaska has more to do with the fact that it mostly wasn't scraped clean by glaciers in the last ice age. http://mapsof.net/map/north-america-glaciation-map#.U-rZu0jz2B0
[PS] Fixed link. Please learn how to do this yourself with the link button in the editor.
wili @57, the idea that soils under previous continental ice sheets are poor because the area was "scraped clean" or "scraped back to bedrock" does not stand up to scrutiny. In particular, the key feature of continental ice sheets is that they are formed in situ by snow failing to melt in the summer. There is some motion of the ice, due to pancaking as areas of higher snow deposition build up to great altitudes but that motion is limited in extent. Areas directly under centers of motion will experience almost not lateral movement. At the edge of the ice sheet, lateral movement will be greater but not scour deep as the ice sheet will only be tens of meter, not thousands of meters thick at those locations. (The situation is quite different, of course, for mountain glaciers.)
Further, the notion that the regions of poor soil in Canada were "scraped clean" does not stand up to compasison of soil types. The very good soils of the Canadian Prairie were covered by a continental ice sheet up until eleven thousand years ago, with the best soil covered the longest:
No Canadian soils on the mainland were covered by ice sheets after eight thousand years ago, so if that were the limiting factor there has been plenty of time to reestablish soils of some variety, even if not suitable agriculture. Further, if "being scraped back to the bedrock" were the problem, regosols would be common throuhout Canada. Instead they are rare. They are dominant in only limited regions (as can be seen by playing around with the Soils of Canda map.
Consequently, I would conclude that the major factors limiting soil quality in Candada have been large areas of poor drainage, particularly south of Hudson bay, large regions were cold climates make them unsuitable for grasslands but suitable for forests, and even larger regions in which the formation of permafrost limits the development of soils.
Finally, glaciation was very extensive in Alaska in the last glaciation, albeit not in the Yukon valley and delta formation.
Ma Rodger@ 54.... those figures were not mine. They were from Scaddenps references I believe. There are different deficiencies that can add to a soil not being suitable for agriculture. .. some like pH can be remedied very quickly. I believe we should stop using fossil fuels not because of the CO2 as much as the fact that we are using up a resource that is not going to last forever.
CB Dunkerson.... the overwhelming evidence is all tied to one fact that CO2 drives global temperature. ... which I believe it does. ... however maybe on a much smaller scale than is hypothesized. Ten years ago I expected much more warming than we have seen despite continuing to dump record amounts of CO2 into the atmosphere. How do you justify that discrepancy in your mind?
[TD] That topic is relevant to the Models Are Unreliable thread. Take the conversation there. Or to the Increasing CO2 Has Little to No Effect thread
Donny, there is no discrepancy in need of justification. Warming over the past ten years has been in line with predictions.
I suspect you are talking about 'lower atmosphere warming', which has been relatively low the past ten years. However, as that represents only about 2% of the total warming it isn't remotely representative. If you look at the full picture (mostly the oceans) then warming has continued unabated. You can get basic info on this at;
http://www.skepticalscience.com/global-warming-stopped-in-1998.htm
[TD] Donny, if you want to respond to this comment, do so on that other thread that CB linked. Everyone else, likewise.
Thanks TomC. But do you have some sources to back up your claim? I know it's just a wiki site, but these are the passages from there that reinforced my crusty memory of hearing this explanation (see link above):
So if you have specific studies that speak against these claims, I would be most interested. Your observations are interesting, but there could be various other reasons for the distributions of soil you mention.
...Mod wrote "we wish" ;-D
Tom Curtis @56:
Thanks for digging out those additional links. Much of the information is consistent with my understanding of the geology - although I didn't download the 280MB PDF, and couldn't get a response from the matching web page.
I know there are areas of potential - I mentioned the Peace River area (N. British Columbia and Alberta) and Clay Belt (N. Ontario) earlier, and both show up in the map you provided.
As for Mills: it's not so much that I'm convinced he's wrong - he just hasn't convinced me he's right. On the basis of what is in that paper, I think the evidence is insufficient.
Tom@58, Will@62:
The geology of the last glaciation in northern Canada is rather complex, and was much-misunderstood until the late 1970s/early 1980s. There was no large single ice cap, and ice on the west side of Hudson Bay did not move out from the Bay - it moved in.
This paper does a good job of summarizing the history, and is just one of many sources of information I turned up using this Google search. Before Tom brings up my non-expert status, I'll admit to a conflict of interest: I participated in some of the relevant geological field work in the Keewatin District in 1977, when working for Bill Shilts (whose name pops up in the references of the paper linked above).
As for soil development: I have pictures (non-digital!) from the Keewatin district where it looks like the glaciers left just a short while ago. Soil development on those deposits is very, very slow in current climates. (The Keewatin District, on the western side of Husdon Bay, above 60N, is not really on our list of potential agricultural zones, though - by the time it is warm enough for agriculture, we'll all be screwed anyway.)
Continental glaciers do extensive scouring in their interiors, and deposit much glacial till in their outer regions. As they retreat, you'll expose a wide variety of till deposits, moraines, course fluvial deposits, and lake deposits where melt water is blocked from draining. Much of the final surface deposits will be the result of that final retreat stage. Many of those deposits make pretty poor soils.
In western Canada, natural drainage to the north was blocked by the retreating ice, and large glacial lakes formed. Glacial Lake Agassiz covered much of Manitoba, and Glacial Lake Regina covered a large area in southern Saskatchewan. Both created clay deposits which have a large effect on soil characteristics in those areas (as well as making the terrain as flat as last night's beer). The Clay Belt in northern Ontario also originates in Glacial Lake Ojibway. Even the history of the Great Lakes is strongly affected by the drainage patterns associated with glacial retreat.
To keep this all on topic - there is much to be considered when examining the suitability of an area for agriculture, under shifting climates. Lack of suitable soil is a serious issue (although all my direct personal experience is from a Canadian outlook).
Donny @59.
You continue you troll-like interventions here.
You say "those figures were not mine" yet you were telling us @50 "so let's discuss what we have learned." You now disown your own interpretation of the learning you previously accepted and adopted. Most informative.
Now when it comes to soil development, it would be good to hear the wise words of somebody qualified in perhaps Environmental Biology. Yet despite you claiming to be such a one, you then tell us "There are different deficiencies that can add to a soil not being suitable for agriculture. .. some like pH can be remedied very quickly." (Unedited quote) So a question - in referring to an exemplar deficiency caused by pH that "can be remedied very quickly," do you refer to soil acidification or soil acidity? And another - how does this fit within the "different deficiencies" present in the soils of Northern Canada discussed here? You tell us "some ... can be remedied very quickly" but what of the important ones, the ones with non-trivial solutions, deficiencies that presumably if not "remedied" will do more than "add to a soil not being suitable for agriculture," the ones that mean farming cannot begin at all, deficiencies like zero soil depth.
Of couse you may feel your expertise does not stretch far enough to make a useful judgement on these matters. But, hey, here is your chance to show us what you're made of.
Thanks, Bob.
Ma... I'm not sure if you missed some posts. ... since some were rerouted to more appropriate thread. ... but I was being told that the southern Canadian soil was no good because it had been scraped clean by glaciers. I was also told it would take millions of years for the soil to recover. .. and I was given a reading assignment. After doing my assignment the studies found that the soil in surrounding areas took 8000 years to recover. And since the glaciers left that area 12000 years ago it was backing up what I had said earlier. <Snip> Now for your second paragraph. ...
[PS] on very thin ice.
Before this discussion gets completely derailed, the original bone of contention is whether new lands, rather than existing suitable land, can be opened for agriculture due to an improved climate closer to the poles, at the same rate as other lands are lost.
I dont see the revelance of papers on existing forest conversion where its suitability already existed. I can find no suggestion in the paper that conversion of boreal forest to agriculture is due improved climate. The paper states that highest rates of conversion were just after WW2 when climate was colder.
Just to put the "new agricultural lands" idea into perspective, and in lieu of boots on the ground, compare the following views from google Earth:
Prime agricultural land north of Saskatoon
This image is centered on Warman. To get the relevant view, tilt twice so that you are looking north, then (while retaining Warman in the forground), scale in to 200 m. Having done so you will see the extensive farmland, which has been plowed, and is evidently intended from growing grains. You will also see the system of small lakes that are a consequence of past erosion from the Laurentide ice sheet.
This map will help you place the view, which is evidently looking north toward the black (best) soil, although whether it is in that belt or just below it I am unsure.
Scrolling the view north you find greener fields, and then forest, but throughout the pattern of extensive lakes persists. Therefore the glacial erosion pattern certainly persists through the best soil of the Canadian Prairie.
Non-Agricultural Land north of Preston Lake
Looking further north within Saskatchewan, we have the view from Preston Lake. Again, tilt twice to look north but focus in till the scale is 500 meters rather than 200 meters, again retaining the main feature (Preston Lake) in the foreground. We focus in to 500 meters because the 200 meter resolution is not available in this part of Canada.
It is interesting to see that here we have a similar pattern of lakes formed by glacial erosion to that found in south Saskatchewan north of Saskatoon. Here, however, the green represents forests. A large proportion of the dry land, however, is taken up by of-white formations that apparently resist tree cover. Comparison with the far whiter sands near lake Athabasca (further north) suggest the of-white formations are not sand, but rock. Indeed, a description of a minereological claim in the area states:
The of-white exposed portions are, therefore, most probabily the pre-cambrian shield rocks as shown here:
Or possibly exposed areas of the underlying sandstone/limestone as shown here:
Clearly this area, while apparently very suitable for a uranium mine, is not at all suitable for agriculture. The brunisolic soils of the area apparently provide a thin cover over the base sedimentary rocks, with earlier (and harder) exposed pre-cambrian rock covering much of the territory. The soil under actual forested areas is probably thicker than in the photo above (where the clearing is probably a clearing for a reason), but is not thick.
Non-agricultural land in Northern Quebec
Finally, for completeness, is the area around Lac Bienville in northern Quebec. Again, centering the google Earth image on the lake, tilting twice to look north and and focusing in to the highest resolution view (500 meters), we see a similar, but for more intensive glacial erosion pattern. Further, as with north Saskatchewan we see extensive pale regions marking regions, whose nature can be determined from this surface photo taken slightly further north at Bienville Sud:
I do not feel the need to point out why that land is not suitable for agriculture.
I'll make an important caveatte that google Earth and photos found on the internet are no substitute for on ground experience, so that my discussion below is premised on the supposition that anybody with onground experience in these regions agrees with my assessment of them. (In this discussion, by "anyone" I primarilly mean Bob Loblaw, who has previously asserted such experience.)
I think this comparison of agricultural vs non-agricultural shows very clearly why land in northern Canada is for the most part unsuitable for agriculture quite independent of current climate. The past history of climate in the region is relevant because formation of soil would have been far faster in a tropical region (for example), but that is no help over the next few hundred years. Even such soil that does exist in northern Saskatchewan is brunisolic, ie, it is "... a stage in an evolutionary sequence that begins with an unweathered parent material (Regosolic soils) and ends with development of a “mature” forested soil of the Podzolic or Luvisolic orders" and as such is unsuitable for agriculture in any event.
Finally, scaddenp @68 mentions Hobson et al (2002). That paper discusses the conversion of forest to agriculture in a belt across the middle of Saskatchewan, more or less on the latitude of Pince Albert, and hence mostly south of Edmonton. As such, it is mostly in, or on the northern edge of the black soil region shown in the map above (first image). Clearly statistics on clearing forest in regions of known good agricultural soil have no bearing on the potential for conversion from forest to agriculture further north where there are large areas of exposed pre-cambrian rock and the soils are of a much lower quality.
As a footnote, I have found a new online map of canadian soils which has the distinction of also being able to show simultaniously the limits of current agriculture. That feature helps show the importance of soil type in placing the northern limit of agriculture. It shows the limited conversion of luvisolic soils to agriculture in Alberta and Saskatchawan, along with the almost complete absence of conversion of brunisolic soils. In west Ontario, regions with brunisolic soils on the US border are still agriculture free. Anybody arguing the brunisolic soils of northen Aberta and Saskatchewan will suddenly become suitable for agriculture due to climate change need first to explain why equivalent soils on the US border are currently unsuitable for agriculture.
Further to my post @69, here is a map of the geological regions of Canada:
It is interesting to note that Canadian agriculture is largely confined to the spur in Ontarion between Lake Huron and Lake Erie, and to the Interior Platform in Manitoba, Saskatchewan, Alberta and British Columbia.
www.pbs.org/newshour/making-sense/why-does-the-govt-pay-farmers/
Ending CRP would help in the extremely unlikely event that there was a shortage.
[PS] This is not science. It is not even basically relevant. Other commentators have produced data to support their argument that climate change is not going create new farmland due to warming temperatures at a rate commeasurate with crop decline in other sectors. Things like actual studies, soil maps etc. Change of land use in existing areas of arable soils are not relevant. Expect further offtopic, essentially political points to be deleted.
Donny @71, the total cost of the Conservation Reserve Program from 1995-2012 was 31.5 billion dollars, for an average payment of 34 thousand dollars per recipient, or 1.9 thousand dollars per annum per recipient.
For comparison, farm commodity subsidies amounted to 177.6 billion dollars over the same period, for an average of 60 thousand dollars per recipient, or 3,356 per annum. Further, there were only 1.4 million eligible recipients for the CRP, compared to the 2.9 million recipients of commodity subsidies. In all, the CRP represented only 13.3% of US farm subsidies. The 1.75 billion payed out in CRP in 2012 represented just 0.4% of gross farm income in the US, and 1.3% of net income.
Given these statistics, the claims made about the CRP at the site to which you link are hardly credible. Having acreage under cultivation results in higher net subisidies for the farmer, which are more easilly obtained. Once income from sale of products is included, it can only be commerically advantagious to have acreage rented under the CRP program if that acreage generates a marginal return in the first event. Further, the CRP does result in a net conservation gain, so that its stated purpose is its most likely actual purpose. (Arguing that a minor tie up of land focussed on degraded land that results in improvement of land quality and recovery in population of threatened species was not introduced for the stated purpose of conservation looks very like a conspiracy theory to me.)
Finally, in total, as of 2014, 5.62 million acres of farmland were tied up by CRP contracts. That is just 0.55% of total US agricultural land. So even it, as per your fantasy, the land tied up by the CRP was as productive as the rest of US agricultural land, releasing it would not compensate other than to a minimal extent to the expected loss in agricultural productivity from global warming.
To the moderator, I know I probably provoked him by posting a sensible post on a related topic, such that Donny had to rush in to fill the space with distracting nonsense, but surely Donny's quota of irrational, unsupported online "arguments" is used up. If he cannot state a cogent case, with clearly linked supporting evidence, why is he permitted to waste our time?
[JH] As long as you and others choose to respond to Donny's "distracting nonsense" before a Moderator can take action to delete his post, the problem will persist.
Tom:
Thanks for the thorough seach for resources on Canadian physical geography and soils. I agree with your assessment of the various non-climate limitations to agriculture in the parts of Canada you cover.
Although you avoid the use of the term "expertise" (thanks!), my training is in physcial geography, and my "boots on the ground" experience includes being born in Manitoba, growing up in southern Ontario, doing geological field work in Nunavut (formerly Northwest Territories, along the western shores of Hudson Bay) and Ontario, climatological field work around Churchil, Manitoba, and 25 years of living in Alberta and Saskatchewan. So I have seen may of the areas under discussion. I travelled to every province and territory in Canada, so my boots are rather worn.
My personal experiences still do remain somewhat anecdotal, but I can comment further on some of the information sources you have uncovered. Your last link to the soils map is most interesting - the web page also provides maps of such things as "local surface form" and "local drainage".
The surface form map indicates that the areas north of the current agricultural zones are often "hummocky (or irregular)". Much of this area (frequently coinciding with the "Canadian Shield" portion of the geological map you provide) is definitely less suitable for agriculture based on changing slope - hills, valleys, lakes, exposed rock, etc. Things get flatter again in the "green zone" in the middle of the Shield area on the map, but that area (the Hudson's Bay Lowlands) is generally poorly drained - which can be see on the "local drainage" option of the soils map referred to previously.
An additional source of more complete (i.e., non-anecdotal) information on the landscape can be found at the National Atlas of Canada's Toporama web site, which provides an interactve topographic mapping tool. With this tool you can zoom in on nearly any part of the country. I'll show three here, but following the link to the site is worth the trip.
You mentioned the area north of Prince Albert, in Saskatchewan. I have worked in that area. Here is Toporama's view of the agricultural fringe in that area:
The southern protion of the map is in an agricultural area. Note the many roads, lack of green shading (forest), and lack of contour lines - this is agricultural land. To the north, where the green indicates forest, note that roads are now lacking (and the ones that are there cannot follow a straight path), and we're starting to see a lot of lakes and areas marked as swamp. Not exactly an ideal candidate for agricultural expansion.
The second map I'll show is north and east of Quebec City, in eastern Canada. Although the St. Lawrence River lowlands are excellent agricultural land, we see again that the land on the northwest part of the map is severely limited by terrain. Lovely scenery, that Canadian Shield, but not prime farming land no matter what the climate.
And last, a map a couple of hours north of Toronto, Ontario. We're now in the southern-most part of Canada, where even current climate is amongst the best in the country, but again we run into the Canadian Shield and its irregular terrain, lakes, swamps, etc.
Same situation: white areas to the south indicate agricultural zones, but forested areas to the north of the map are "cottage country" - Toronto's weekend and summer playground.
As a last note, I have decided to not attempt to find photos on the Internet, but I will promise to try to fire up my old computer and scanner to scan a few of my personal phots from some of these areas.