Category: GIS

Posted on

The climate–demography vulnerability index of my mother-in-law

ResearchBlogging.orgAnother dispatch from the outer reaches of GISland. Yesterday’s post on the likely consequences of climate change around my mother-in-law’s farm in Kenya got me thinking that it would be nice to see where that locality fits in the global vulnerability scene. One can actually do that thanks to a recent paper in Global Ecology and Biogeography. ((Samson, J., Berteaux, D., McGill, B., & Humphries, M. (2011). Geographic disparities and moral hazards in the predicted impacts of climate change on human populations Global Ecology and Biogeography DOI: 10.1111/j.1466-8238.2010.00632.x))

The authors start by calculating something they call Global Climate Vulnerability Index (CVI)

…by combining climate change forecasts with current relationships between human density and climate. We further refined the CVI by contrasting predicted vulnerabilities with demographic growth rates to create a climate–demography vulnerability index (CDVI) reflecting the spatial disparities between demographic trends and climate-consistent population growth.

The global map of CDVI is Fig. 5 in the paper. But how to get that into Google Earth? Thanks to the raw GIS files from one of the authors, and some R magic from friend and occasional contributor Robert, I now have a kmz file of CDVI, on top of which I can easily plot the location of the mother-in-law’s spread at Gataka near Limuru. In the map below, dark blue is bad, light blue less so.

Gataka turns out to have a slightly positive CDVI.

Highly negative values [of CDVI] … represent low-vulnerability situations where current demographic growth is much lower than climate-consistent population growth, while highly positive values … represent high-vulnerability situations where current demographic growth vastly exceeds climate-consistent population growth.

So, bad news for the mother-in-law, but not actually as bad as I feared. I wonder if I can persuade her to diversify. Perhaps into indigenous leafy greens. And sorghum, as maize seems to be heading for trouble. SPAM says sorghum should be the main crop here anyway. It may yet turn out to be right.

Posted on

More messing about with Droppr

I continued my exploration of IFPRI’s wonderful Droppr software by looking at its future climate tool. You click on a spot on the Earth and it tells you how total precipitation and average temperature will change, for each month of the year. Again, I did it for the mother-in-law’s farm, and this is the result:

Looks pretty bad, at least for temperature. Although of course, for maize at least, which is the main food crop in that area, what you really want to know is peak, rather than average, temperatures. That’s according to a study by David Lobell and Marianne Bänziger we nibbled a few days back, and which recently got a big write up in The Economist:

Days above 30°C are particularly damaging. In otherwise normal conditions, every day the temperature is over this threshold diminishes yields by at least 1%. Moreover, days where the temperature exceeds 32°C do twice the harm of those at 31°C. And during a drought, things are worse still. Then, yields take a hit of 1.7% per day over 30°C.

Posted on

Ground-truthing SPAM

Jeff Horwich has an interesting post over at HarvestChoice Labs looking at the effect of the tsunami on agriculture in northern Japan. He used the Droppr tool, which combines Google Maps with lots of other data, in this particular case the world-wide crop distribution data from the Spatial Production Allocation Model (SPAM). In SPAM

…tabular crop production statistics are blended judiciously with an array of other secondary data to assess the production of specific crops within individual ‘pixels’ – typically 25–100 square kilometers in size. The information utilized includes crop production statistics, farming system characteristics, satellite-derived land cover data, biophysical crop suitability assessments, and population density.

Intrigued, I decided to do a little lighthearted ground-truthing of the SPAM data. I only looked at one location, I admit, but what I found was a bit disappointing. I zoomed in on the location of the mother-in-law’s farm in the Limuru highlands (-1° 4′ 39.60″, +36° 40′ 44.80″). Here’s what the place looks like.

This is the view from space, courtesy of Google Earth. ((The blue dot is directly over the house you see in the middle of the photograph above, which was taken from the S.))

Now, according to the SPAM methodology that’s a cropping intensity of 0.59%, with the main crops being sorghum, sweet potato/yam, groundnut, banana/plantain, potato, coffee and sugarcane.

Jeff very niftily embedded both a map and a spreadsheet of the production data for the main crops in his post, but I wasn’t able to work out how to do that. So you’ll have to make do with this wholly inadequate screengrab, I’m afraid.

What my mother-in-law and her neighbours actually grow is maize, beans, potato and tea, tea and tea. A pretty different set of agrobiodiversity to what SPAM thinks. And I think she would be surprised at the low value of cropping intensity. This is a very high-potential area.

Anyway, that’s only one data point. It would be interesting to know from the SPAM guys if there’s a more systematic attempt going on to check on, and refine, the results of the model.

Posted on

Dams, lying links, and databases

A post on e-agriculture about information resources related to water in agriculture allows me to update, on the occasion of World Water Day, a piece we had here some years back. The links in that old post of ours no longer point to the things they used to, but if readers are still interested in that African dams database they can now find it elsewhere. Alas, I couldn’t get the Google Earth file to work, but if you do the work-around from the Excel file you get the map below. I’ll reiterate my original questions, to which I have no better answer now than four years ago, alas:

I would guess that the effect of dams and new irrigation schemes on local wild biodiversity is usually negative, but is that necessarily always the case also for agro-biodiversity? I suspect so, but is there a possibility that at least sometimes existing crop genetic diversity is simply displaced a bit geographically or ecologically within the same general area and augmented by new crop genetic diversity adapted to the new conditions?

Posted on

Training manual for GIS analysis of agrobiodiversity data

Great to see “Training Manual on Spatial Analysis of Plant Diversity and Distribution” finally out, courtesy of Bioversity International. Well worth the wait, and not just because I get called a pioneer in it. Congratulations to Xavier Scheldeman and Maarten van Zonneveld for addressing a very important need.

This manual has been published as a result of the increasing number of requests received by Bioversity International for capacity building on the spatial analysis of biodiversity data. The authors have developed a set of step-by-step instructions, accompanied by a series of analyses, based on free and publically available software: DIVA-GIS, a GIS programme specifically designed to undertake spatial diversity analysis; and Maxent, a species distribution modelling programme. The manual does not aim to illustrate the use of each individual DIVA-GIS and Maxent command/option, but focuses on using GIS tools to help answer common questions relating to the spatial analysis of biodiversity data. Throughout the manual, the importance of proper sampling is stressed; however, it is beyond the scope of the document to elaborate on sampling theories. The manual also does not discuss the statistical analysis of diversity data in detail; instead, when statistical methods and programmes are mentioned in the text, the reader is referred to alternative reference materials for further information.