were collected and visualized in GIS to show the energy use distribution, as in Fig 2a. Also the modeling
could estimate monthly and hourly building energy use to better understand the energy use dynamics.
The validation using a building energy and water use dataset in Manhattan in 2012 provided under the
Local Law 84 (LL84) [24] shows a considerable accuracy, with the Coefficient of Variance Root Mean
Square Error of 0.69, less than 5 times of the strict ASHRAE standard for a single building [25].