WCS (Web Coverage Service) data access to SoilGrids

[Tomislav Hengl]

You can also download SoilGrids data by using WCS (Web Coverage Service).

The easiest thing to do is to use QGIS to connect to WCS:

- Add layer from WCS -> http://webservices.isric.org/geoserver/wcs?

Then from QGIS (or which ever desktop GIS software you use) you can
download any data in any format you prefer.

An alternative is to script (always a better path). SoilGrids tutorial
for R is available at:

http://gsif.isric.org/doku.php/wiki:tutorial_soilgrids#wcs_data_access

good luck with your work,

T. (Tom) Hengl
Researcher @ ISRIC - World Soil Information
Url: http://www.wageningenur.nl/en/Persons/dr.-T-Tom-Hengl.htm
Network: http://profiles.google.com/tom.hengl
Publications: http://scholar.google.com/citations?user=2oYU7S8AAAAJ
ORCID ID: http://orcid.org/0000-0002-9921-5129

On 06-01-17 06:43, John Mendelsohn (RAISON) wrote:
> Dear Tom
>
> Many thanks for your very quick and helpful response. I looked around at the new datasets
> and see that it is possible to download blocks of data, each block covering something like
> 200 x 200 kilometres in small files. And then there is the site which allows downloads of full
> sets of data for each variable. Those files are really huge and would take ages to download
> given the internet speeds we have here. Is there something in between, for example that
> allows downloads of data for a whole country, or in bigger blocks? I searched and searched
> but could not see such a facility.
>
> With kind regards - John
>
>
>
> Subject: Re: ISRIC data sets
> To: "John Mendelsohn (RAISON)" <jo...@raison.com.na>
> Copies to: stephie mendelsohn <mendelsoh...@gmail.com>
> From: Tomislav Hengl <tom....@isric.org>
> Date sent: Thu, 5 Jan 2017 12:48:49 +0100
>
>>
>> Hi John,
>>
>> Are you using the most up-to-date data?
>>
>> We now also predict values for arid zones:
>>
>> https://www.soilgrids.org/#/?lon=13.754882812499998&lat=-19.7667035517
>> 16972&zoom=6&layer=geonode:phihox_m_sl2_250m
>>
>> See also:
>>
>> http://gsif.isric.org/lib/exe/fetch.php?media=wiki:soilgrids250m_globa
>> l_gridded_preprint.pdf
>>
>> RE "I noted that you have very large numbers of soil sampling
>> locations in Angola, and that immediately suggested to me that the
>> grid data should be rather good, or at least considerably better than
>> grid data for areas which have been little sampled.Would my assumption
>> about Angola's grid data be correct?"
>>
>> Not necessarily. In regression, it is more important how
>> representative are the points of the feature space and how accurate is
>> sampling/measurement of the ground data, than what is the data volume.
>> I could probably produce better SoilGrids with 2-5 times less points
>> than what I have at the moment but if:
>>
>> 1. Points were sampled using objective sampling plan such as latin
>> hypercube or similar, 2. If there were no harmonization issues (no
>> bias / minimum possible noise), 3. If all points refer to the same
>> time period. Same support size etc.
>>
>> But yes if there are more points - the models do also get better.
>>
>> Please subscribe to our mailing list at:
>>
>> https://groups.google.com/forum/#!forum/global-soil-information
>>
>> T. (Tom) Hengl
>> Researcher @ ISRIC - World Soil Information
>> Url: http://www.wageningenur.nl/en/Persons/dr.-T-Tom-Hengl.htm
>> Network: http://profiles.google.com/tom.hengl
>> Publications: http://scholar.google.com/citations?user=2oYU7S8AAAAJ
>> ORCID ID: http://orcid.org/0000-0002-9921-5129
>>
>> On 05-01-17 11:30, John Mendelsohn (RAISON) wrote:
>>> Dear Tom Hengl
>>>
>>> Greetings from Namibia. We have made great use of the soil grids you
>>> helped compile in a number of profiles and atlasses. I just noticed
>>> that your organisation has an email address (so...@isric.org) to
>>> which copies of publications should be sent, and I will send copies
>>> immediately via wetransfer. Let me know if you would like to have
>>> copies sent to you as well.
>>>
>>> My reason for writing is twofold. First, is to ask why data for your
>>> various parameters are missing for coastal areas in southwestern
>>> Angola and western Namibia? Would this be because these areas are
>>> just too arid for most plant growth, or because the soils are often
>>> very saline or sandy? Any information on the reasons for these gaps
>>> would be much appreciated.
>>>
>>> Much of my work is now in Angola and that leads to a second
>>> question, which concerns confidence in the soil grid parameters for
>>> Angola. I noted that you have very large numbers of soil sampling
>>> locations in Angola, and that immediately suggested to me that the
>>> grid data should be rather good, or at least considerably better
>>> than grid data for areas which have been little sampled. Would my
>>> assumption about Angola's grid data be correct? Your comments would
>>> be very useful to me and others who use your data sets.
>>>
>>> I look forward to your ideas and response, and also wish you all the
>>> best for 2017.
>>>
>>> John Mendelsohn
>>> RAISON
>>> PO Box 1405
>>> Windhoek, Namibia
>>> Telephone Namibia +264 81 148 2385; Angola +244 938 248 267
>>>
>>> www.raison.com.na
>>> www.the-eis.com - the one-stop-shop for environmental information in
>>> Namibia
>>>
>>>
>
>
>
> John Mendelsohn
> RAISON
> PO Box 1405
> Windhoek, Namibia
> Telephone Namibia +264 81 148 2385; Angola +244 938 248 267
>
> www.raison.com.na
> www.the-eis.com - the one-stop-shop for environmental information in Namibia
>
>