Combined Datasets

A harmonised, high-resolution (0.01°) surface NO2 dataset from the University of Leicester is presented in this section. Surface concentrations have been calculated by combining in-situ measurements with OMI tropospheric columns as part of a Land Use Regression model (see Panda deliverable report D2.2). Data is presented for the North China Plain (ppb) and the Pearl River Delta (μgm-3) regions for January and July 2010/2013, and for Hong Kong SAR for the period 2005 to 2015. Fine-scale information has been derived by correlating the in-situ data with known covariates from other datasets (e.g. urban area coverage from MODIS and OMI, surface elevation from ASTER, and road networks from OpenStreetMap). The Land Use Regression model was validated using in-situ data sampled by 5-fold cross-validation; overall, the modelled data shows good agreement with surface concentration measurements over both regions (R2 = 0.54 and 0.70 for the North China Plain and Pearl River Delta, respectively). However, insufficient OMI coverage from excessive cloud cover and degradation caused by the row anomaly has caused nonphysical artifacts to appear due to sampling biases.

Data files can be downloaded on netCDF format from the list below.

North China Plain

Pear River Delta

Hong Kong SAR

  • 2005-2015 (zip Archive containing files of monthly averages)