Water Use Vector Discrepancy from Feb. 3, 2010 to March 10, 2010.
Short Version: The water use vector posted as of February 3, 2010 for the 2002 US Producer Price Model had errors from the original posting which were discovered on March 10, 2010. If you used the model during this period for water withdrawals, please re-do your estimates with the current model. The water use vector only affects water withdrawal estimates, so other model results (such as economic activity, greenhouse gas emissions, etc.) do not have to be re-calculated and are unchanged.
Long version: The methodology used to estimate a vector of water withdrawals is documented in the peer-reviewed article below:
Direct and Indirect Water Withdrawals for U.S. Industrial Sectors, By Michael Blackhurst, Chris Hendrickson, Jordi Sels i Vidal, Environmental Science & Technology 2010 44 (6), pp. 2126-2130.
The estimates use a variety of methods to disaggregate water use by sector within broad USGS water withdrawal categories.
This water use vector is only available for the 2002 US producer price model. The 1992 US model also has a water use vector, but this is based on census data which is no longer available.
Unfortunately, an incorrect dataset was originally posted on the www.eiolca.net
site. This data had several sectors with too much water use (such as the paint sector). When we realized the error, we replaced the vector with the correct one. Model runs since March 10, 2010 should be identical since no other changes have been made. Water use model runs made between February 3, 2010 and March 10, 2010 should be re-done. The water use vector is only used to estimate water withdrawals by sector within the model, so no other results are affected by the incorrect vector.
The correct vector has always been posted on the ES&T website as supplemental information for the article cited above as well as being available for calculations at the www.eiolca.net
site. Only the online data was incorrect, for one month.
We are sorry for this error.