Downloading data from SoilsGrid
400 views
Skip to first unread message
J McArthur
Dec 11, 2017, 7:39:41 AM12/11/17
to Global Soil Information
Dear Dr. Hengl,
Arsenic pollution of groundwater is a worldwide problem.
In the paper sent to your e-mail address, your SoilsGrid data is used to assess the probability of arsenic pollution of groundwater across Pakistan.
The paper has some oddities that I would like to assess, owing to the considerable implications of the paper's findings.
To assess the paper's findings, I need soil-pH at 0 and 200 cm depth for 1200 locations in Pakistan.
Having spent some time attempting to obtain this data through a number of locations in SoilsGrid, I have to admit defeat.
As a simple geochemist, I never use GIS, nor do I program, and most of the explanations of how to obtain data from Soilsgrid are opaque to me - I simply don't understand the language.
Three options are therefore available to me:
1] plug each GPS co-ordinate I have into SoilsGrid and copy the data that comes up and do that 1200 times.
2] spend open-ended time trying to work out how to get the data from SoilsGrid.
3] Ask you how to do it.
Please forgive me for starting this week with No 3, after a frustrating weekend on 1 and 2.
Can you help?
Yours sincerely,
John McArthur
Arsenic pollution of groundwater is a worldwide problem.
In the paper sent to your e-mail address, your SoilsGrid data is used to assess the probability of arsenic pollution of groundwater across Pakistan.
The paper has some oddities that I would like to assess, owing to the considerable implications of the paper's findings.
To assess the paper's findings, I need soil-pH at 0 and 200 cm depth for 1200 locations in Pakistan.
Having spent some time attempting to obtain this data through a number of locations in SoilsGrid, I have to admit defeat.
As a simple geochemist, I never use GIS, nor do I program, and most of the explanations of how to obtain data from Soilsgrid are opaque to me - I simply don't understand the language.
Three options are therefore available to me:
1] plug each GPS co-ordinate I have into SoilsGrid and copy the data that comes up and do that 1200 times.
2] spend open-ended time trying to work out how to get the data from SoilsGrid.
3] Ask you how to do it.
Please forgive me for starting this week with No 3, after a frustrating weekend on 1 and 2.
Can you help?
Yours sincerely,
John McArthur
maria.ruiperezgonzalez
Dec 18, 2017, 10:43:37 AM12/18/17
to Global Soil Information
Dear John,
If I understood you well you want to know the value for pH predicted under SoilGrids approach for 1200 locations in your country, right?.
For that you should follow this steps, as suggestion:
step1: go to https://www.soilgrids.org and download the 7 layers of pH at 1 km resolution from your country.
step2: derived the weighted average of the pH predictions within your target depth interval (0 -200 cm) the trapezoidal rule described in:
step3: do an overlay of your points (the 1200 locations) with the aggregated raster output of step 2. Then you will have the pH predicted value for interval depht of 0 -200 cm for your target locations.
I am afraid for doing these operations you will need to use GIS software like QGIS or R studio, for example.
Hoping this helps,
Kind regards,
María
Tomislav Hengl
Dec 19, 2017, 7:31:49 AM12/19/17
to Global Soil Information
Another opportunity to get values of soil properties for Pakistan is to use the REST SoilGrids service. This is an example of output for soil pH in H2O for a point in Pakistan:
which visually looks like this:
If you use R, you can fetch values for points automatically using:
This function is however NOT recommended for large point data sets (>>100).
Yes for further aggregation please use the trapezoidal rule.
On Monday, December 11, 2017 at 1:39:41 PM UTC+1, J McArthur wrote:
ahmedatia80
Jul 21, 2020, 10:24:55 AM7/21/20
to SoilGrids Community Forum
Hello,
I try to use the REST.soilGrid function but the over method is returning a data.frame and not SpatialPixelsDataFrame.
R version 3.6.3
> coordinates(pnts) <- ~lon+lat > proj4string(pnts) <- CRS("+proj=longlat +datum=WGS84") > pnts class : SpatialPointsDataFrame features : 2 extent : 5.36, 10.65, 51.48, 51.81 (xmin, xmax, ymin, ymax) crs : +proj=longlat +datum=WGS84 +ellps=WGS84 +towgs84=0,0,0 variables : 1 names : id min values : p1 max values : p2 > ## REST example: > soilgrids <- REST.SoilGrids(c("ORCDRC","PHIHOX")) > ov <- over(soilgrids, pnts) > str(ov) 'data.frame': 2 obs. of 2 variables: $ lon: num 10.65 5.36 $ lat: num 51.8 51.5 > class(ov) [1] "data.frame" |
Muhammad Iqbal
Jul 21, 2020, 12:11:21 PM7/21/20
to SoilGrids Community Forum
Dear John,
This is quite easy in GIS environment. If you can provide co-ordinates for the selected points, I can do it for you.
Regards
Dr. Muhammad Iqbal
This is quite easy in GIS environment. If you can provide co-ordinates for the selected points, I can do it for you.
Regards
Dr. Muhammad Iqbal
David Gibbs
Jul 23, 2020, 11:58:01 AM7/23/20
to SoilGrids Community Forum
Hello SoilGrids community,
I want to use the "most probable" soil type rasters found at https://files.isric.org/soilgrids/latest/data/wrb/MostProbable/. Where can I find a key to interpret the values in those rasters? Thanks very much.
Sincerely,
David Gibbs
Muhammad Iqbal
Jul 23, 2020, 1:37:38 PM7/23/20
to SoilGrids Community Forum
Dear David,
I think you can find the associated labels in the .xml or .qml files in the link below;
Regards
Reply all
Reply to author
Forward
0 new messages