★★★★☆

• Intro

  • Context

    Atmospheric water vapor → a key role in the global water and energy cycles

    Accurate estimation → critical to climate analysis and model validation

  • Objectives

    (i) to evaluate absolute and relative accuracy of the PWV products at a pixel-wise level in reference to in situ measures of radiosonde and sunphotometer

    (ii) to discriminate fact and fiction of the long-term trend in the pixel-wise PWV products in a temporal sense

    (iii) to portray spatial and temporal distribution of global water vapor for the study period using the most reliable products.

• M&M

  • Data

    MODIS precipitable water vapor (PWV) products for 2000-2017

    TIR Collection 6 (C006), its updated version Collection 061 (C061), NIR products C061

    Ground observational data from global radiosonde and sunphotometer networks for evaluation

  • Methods

    Linear regression & Mann-Kendall (M-K) analysis → to identify temporal trend or systemic error

• Results & Contributions

  C061_TIR data showed improved accuracy in terms of bias, standard deviation, mean absolute error, root mean square error, and coefficient of determination, regression slope and intercept. (*)

  Among the PWV products, C061_NIR data achieved the best overall performance in accuracy evaluation. (*)

  The C061_NIR revealed the PWV had a multi-year average of 2.50 ± 0.08 cm for the globe, 2.03 ± 0.06 cm for continents, and 2.70 ± 0.09 cm for oceans in 2000–2017. (*)

  The PWV values yielded an increasing rate of 0.015 cm/year for the globe, 0.010 cm/year for continents, and 0.017 cm/year for oceans. (**)

  Nearly 98.95% of the globe showed an increasing trend, 80.74% of statistical significance, mainly distributed within and around the tropical zones. (**)

  Downward trend is an artifact in previous version of thermal-infrared products. (**)

★☆☆☆☆

• Intro

  • Context

    Reliable urban land-use maps → essential for urban analysis

    Very high resolution (VHR) remote sensing imagery interpretation resolved “semantic gap” between the low-level data and the high-level semantic scenes

    Nevertheless, the existing frameworks cannot easily be applied to practical urban analysis because of : 1) the indistinguishable socio-economic attributes of the same ground object layouts; 2) the weak transferability of the supervised frameworks and the time-consuming training sample annotation; 3) the category system inconsistency between the data source and the urban land-use application

  • Objectives

    Achieving an “application gap” breakthrough for urban land-use mapping

• M&M

  A data-driven point, line, and polygon semantic object mapping (PLPSOM) framework (=makes full use of open-source VHR images and multi-source geospatial data)

  Point, line, and polygon semantic objects are represented by the points of interest (POIs), OpenStreetMap (OSM) data, and VHR images corresponding to the scenes in the land-use mapping units, respectively

  OSM line semantic objects are utilized to supply the boundaries of the land-use mapping units for the POIs and VHR images, forming urban land parcels (street blocks)

  To reduce the cost of the data annotation, the training dataset is constructed using multiple open-source data sources

  An enhanced deep adaptation network (EDAN) is then proposed to acquire the categories of the VHR scene images in the case of partial transfer learning

  A rule-based category mapping (RCM) model is applied to integrate the categories of the POIs and VHR images into the urban land-use category system

   

• Results & Contributions

  Proposed method tested in four cities of China

  Achieving a high classification accuracy.

• Statistical analysis of time-series : regression & Mann-Kendall (M-K) analysis 참고