>Contributing Data>Example Uploads>Nitrate and ammonia leaching meta-analysis

Nitrate and ammonia leaching meta-analysis


Dataset

A meta-analysis consolidated data from 91 studies representing 559 observation of nitrate and ammonium leaching under cropland. The dataset includes possible covariates of leaching such as rainfall and crop yield.

Conversion steps

We used the steps as described in the Myanmar example, with the following additional steps:

  • To each Source, we only added the title of the article (bibliography.title) and/or the DOI (bibliography.documentDOI) and let the HESTIA pipeline add the rest of the bibliographic items automatically from the Mendeley library.
  • To each Source, we also added a metaAnalysisBibliography.
  • In the meta-analysis, crop yield was expressed as kilograms of dry matter, however in the HESTIA Glossary crops are expressed in kilograms of marketable weight (e.g., Maize, grain has a default dry matter of 85.8%). Therefore, to each crop Product we added the Property Dry matter from the Glossary with the value 100%. Adding a Property to a blank node overrides the default Property from the Glossary.
  • Some fertilisers were controlled release, or included nitrification or urease inhibitors, and this was specified by adding a Property to the fertiliser.
  • Soil Measurements had defined depth intervals, and these were added using depthUpper and depthLower.
  • Some Measurements had different Sources to the defaultSource of the Site (e.g., a prior study on the same Site may have reported the soil pH), and these were specified for each Measurement, creating a new source.id if necessary.
  • Emissions have required fields for methodModel and methodTier. The method or model used to quantify emissions can change the result and this is a critical field to record, and we identified the correct methodModel in the Glossary for each Emission. Methods and models can also be broadly grouped by their specificity and methodTier and the value for these Emissions is measured.

Validation

The bibliographies for seven studies were not available automatically and these were entered manually. Some coordinates were identified as being in the wrong country or region, and these were corrected (either by referring to the original study if the error was in the meta-analysis, or by using other site location data in the original study instead if the error was there). Some more coordinates were identified as erroneous where they did not intersect with known cropland (according to the MODIS land cover map) and some of these were corrected but others were accepted based on visual inspection of high-resolution satellite images. Some estimates of above ground crop residue were automatically identified as too high, were found to include below ground residue, and were corrected.

Downloads