|Page tools: Print Page Print All|
As part of the Australian Water Resources 2005 (AWR 2005) project, the Australian Bureau of Statistics (ABS) was funded by the National Water Commission (NWC) to model water use data from the Water Account 2004-05 (cat. no. 4610.0) to a regional level; namely, Water Management Areas (WMAs).
This publication presents the methodology used to produce the experimental estimates of regional water consumption for WMAs during 2004-05, the associated explanatory material, and the final regional estimates. The ABS defines 'water consumption' as the sum of Distributed water use, Self-extracted water use and Reuse water use less Water supplied to other users less In-stream use and less Distributed water use by the environment (see Water Account 2004-05 (cat. no. 4610.0) for further details).
The methodology presented here is one of a number of possible approaches to model estimates to a regional level, and was selected as the best compromise based on the available data, resources and time. This release supersedes the Proposed Methodology for Producing Regional Water Use Estimates (cat. no. 4610.0.55.001) release of 18 September 2006.
ABS WATER ACCOUNT 2004-05
The ABS Water Account describes the flow of water from the environment through the economy. The most recent, Water Account 2004-05, was released on 28 November 2006, and presents data for 2004-05. The primary objective of the Water Account is to integrate data from different sources into a consolidated dataset to assist decision-making and inform policy development and direction. The core outputs of the Water Account are water supply and water use tables; collectively called flow tables. The flow tables show the volumes of water supplied to, and used by, the various sectors and industries within each Australian state and territory.
The ABS Water Accounts have been developed using the System of Integrated Environmental and Economic Accounting (SEEA). SEEA complements the System of National Accounts 1993 (SNA 1993). Environmental accounts extend the boundaries of the SNA framework to include environmental resources which occur outside the production and asset boundaries typically measured in the SNA. For further details of SEEA see http://unstats.un.org/unsd/default.htm.
The methods used to compile the ABS Water Accounts, while similar, have varied between each edition (1993-94 to 1996-97, 2000-01 and 2004-05). This is largely as a result of improved reliability and availability of information on the volumes of water flowing through the economy. In compiling the first two editions of the Water Account, the ABS drew on readily available water resource data from various government and non-government organisations, and aggregated these data to produce the flow tables. The process did not duplicate existing data collection activities, but drew together industry, regional and state data into a single system showing the supply and use of water within the Australian economy. In the third edition, more data were sourced from ABS surveys.
DATA SOURCES USED FOR COMPILING THE WATER ACCOUNT
A review of the Water Account 2000-01 highlighted a need for more direct collection of water flow information for input into supply and use tables. The ABS developed a number of purpose-designed surveys (such as the Water Supply Survey, 2004-05); and water-specific questions for incorporation into existing ABS surveys. These surveys include the Agriculture, Mining, Manufacturing, Service Industries Surveys, as well as households surveys conducted in respect of 2004-05.
The use of surveys to collect water supply and use data allows the major water-supplying and water-using businesses and organisations to directly report data to the ABS. Where businesses and organisations did not report water use, estimates of water use were developed using information from like businesses. Direct collection of data simplified the compilation process and reduced the time required to produce the third edition of the Water Account.
ABS data sources that were integrated into the Water Account 2004-05 were the:
The non-ABS data sources that were integrated include:
ABS SURVEY DESIGN AND GEOGRAPHIES
Typically, surveys are designed to produce estimates at the state/territory level or a sub-state geography, such as Statistical Divisions (SD) or Statistical Local Areas (SLA). The level of geography used to design and output ABS surveys are described in the 2001 edition of the Australian Standard Geographical Classification (cat. no. 1216.0).
The 2004-05 ABS surveys that were inputs to the Water Account 2004-05 were designed to produce estimates at the state and SD level. To produce the state- and territory-level tables requires that water use data be classified to the appropriate state or territory where the water is used. Consequently, producing the Water Account 2004-05 required minimal intervention to manually locate the data (i.e. where the water is used) in order to produce the state/territory-level estimates.
USER DEMAND FOR REGIONAL LEVEL ESTIMATES
In addition to state and territory level water use data, there is great demand for small-area water statistics; that is, information for areas smaller than state/territory boundaries. The ABS has recognised the need for water and other environmental data at a lower spatial level than state/territory, and has been exploring options to meet this requirement. As issues of water availability and use are often localised in nature, the development of a methodology for producing small area (or regional) water use statistics has helped meet this demand.
The development of the regional water use methodology builds on other ABS activities designed to improve the availability of regional estimates. Such activities include the development of:
REGIONAL WATER USE ESTIMATES
In contrast to the estimates presented in the Water Account 2004-05, the production of water use estimates at the regional level, in this case the WMA level, required additional intervention to more precisely spatially locate the data. Ideally, to produce detailed regional level data, a census or large sample survey would be designed with the primary objective of producing quality estimates for the specified spatial level, i.e. estimates for each WMA. However, the resources required and reporting burden imposed on respondents would be large. Furthermore, the surveys in respect of the 2004-05 reference year were already in the field when this project was first requested by NWC. As such, no additional data collection was undertaken, and the key focus of the project was developing a suitable method to allocate the state and territory data from the Water Account 2004-05 to the WMAs.
Retrospective spatial placement of data was possible, but is less than ideal since it typically involved modelling data to the required spatial level using a ‘top down’ approach; that is, modelling large area data to smaller areas. Given that direct estimation of total water use (consumption) at the regional level from the 2004-05 survey results was not possible, a methodology that uses the existing survey results and makes use of other existing data was needed.
PRODUCING REGIONAL WATER USE (CONSUMPTION) ESTIMATES
The simplest method of producing regional water use (consumption) estimates is to take known higher-level estimates (e.g. state/territory level estimates) and apportion the estimate according to a known variable (e.g. area or population). When the apportioning method is used, homogeneity or uniformity of the variable to be apportioned, in this case water use (consumption), is required and assumed. Since water use (consumption) across regions is not uniform, an apportionment model will result in estimates with large errors (see Research Paper: A Methodology for Estimating Regional Agriculture Water Use (cat. no. 4616.0.55.001).
Ideally, the methodology applied must allow for regional variability, especially as water use (consumption) can be localised (e.g. mines sites, manufacturing plants etc). A methodology designed around using the 2004-05 ABS survey data that captured major water users within each Australian state and territory was proposed and described in Proposed Methodology for Producing Regional Water Use Estimates (cat. no. 4610.0.55.001).
The final methodology used was:
1. Building a profile of the region for which an estimate is required. In this case, the WMAs formed the basic spatial units and the profile involved establishing the totality of individuals, businesses and organisations within the WMAs. The profile was derived from ABS population statistics, and, the ABS business register and the Australian Business Register. The profile for Australia resulted in spatially identifying approximately 3 million businesses, 20.5 million people and 7.5 million households.
2. For each WMA, individual businesses and organisations within the WMA were identified and then stratified by water-using ‘industry’ and employment size.
3. The surveys used in the production of the Water Account 2004-05 were used to identify the large water-using industries and businesses that needed to be spatially assigned to an appropriate WMA. These surveys include the Water Supply Survey, Manufacturing Survey, Mining Survey, and the Service Industries Survey.
4. Since some of the businesses surveyed operate at more than one site, the data reported was split into major production units, e.g. individual factories, mine sites etc. The water use reported by multi-site businesses was reallocated to each site. For some businesses water use was available for each site. Where water use for multiple sites was not available, an allocation was made based on the size of the production units defined by either production output or employment. It should be noted that the majority of businesses only operate at a single site, or at least within a single WMA. However, these businesses are typically small in size and are not significant water users.
5. After the large water-using businesses were assigned to an appropriate WMA (which accounted for the majority of the water used within each WMA), the water used by the remaining business was derived. The process involved first locating the remaining businesses in WMAs, and then using surveyed businesses to calculate a water use estimate for all non-major water-using businesses within the WMA.
6. A specific methodology was developed to produce the regional estimates for the Agriculture industry. The final method used to produce the agriculture water use estimates involved:
b. Designated irrigation areas as defined in ACLUMP were used to refine the concordance used to allocate crop and pasture water;
c. A basic area concordance refined by excluding land described as urban and parkland was used to allocate stock drinking water;
d. Water use estimates sourced from the 2004-05 Agriculture Survey at the Statistical Division level were allocated to the WMA level; and
e. The data was confronted with the rural Water Supply Survey reported agriculture water use, for distributed water.
7. A specific methodology was developed to produce WMA-level estimates for household water consumption. The estimate includes information from a number of sources including the ABS Water Supply Survey and the ABS Monthly Population Survey. For states where local governments provide water supply data for the household sector, a concordance between Local Government Area (LGA) and WMA was used to allocate data. For other states/territories, estimated population and volume of water supplied for domestic use were used to prorate to WMA.
8. The total water consumption for each WMA were derived from:
b. the estimate for the smaller water users, plus
c. the estimates for any sector specific industries, plus
d. the estimates for any households
9. As part of the standard ABS quality assurance process, the WMA total water use (consumption) estimates were cross-checked with state/territory level estimates from the Water Account 2004-05 and any other known regional sources.
OUTPUTS FROM THE REGIONAL WATER USE (CONSUMPTION) ESTIMATES PROJECT
The aim of the project was to produce water use (consumption) estimates for up to 200 WMAs in Australia, and where possible, provide an indication of the main water use within each WMA. The data is presented in the data cube under the Details tab.
The data presented is:
For some WMAs, a reliable estimate of water use (consumption) was not possible. This can occur because the underlying data is insufficient to produce a reliable estimate where, for example, the WMAs are very small. To overcome this, contiguous WMAs have been merged in order to maximise data output. However, the level of detail finally presented were dependant on the structure of each WMA, the data available for the businesses and organisations occurring in each WMA, and the need to preserve the confidentiality of data providers. Where there is the need to protect the confidentiality of some contributors to specific estimates (i.e. very large water users) these data cells have been suppressed and marked as ‘not for publication’ (n.p.).
DATA QUALITY AND RELIABILITY
The surveys and data used to produce the supply-use tables for the Water Account 2004-05 were purposefully designed to produce high quality state- and territory-level estimates. The regional estimates were derived from these data using the methodology outlined above. As such, the regional estimates were of lower quality and reliability than the state-level estimates presented in the Water Account 2004-05.
Data for the Water Account 2004-05 and subsequent regional estimates of water use are compiled from a range of sources. The water use data provided is of varying degrees of reliability and consistency. While the main data suppliers were asked to indicate the reliability of the data provided, reliability indicators were not provided by all respondents.
In terms of the various data sources, the following applies:
Data from various sources was aggregated at the regional level to form a water use estimate for each region. In general, the regional water use data should be able to be used with a moderate degree of confidence. Where data quality and confidentiality allows, the data for each region was presented by broad categories of use (e.g. agriculture, household, industry), but the degree of confidence in the disaggregated data were lower.
The ABS compiled the regional estimates of water use as part of the Australian Water Resources 2005 (AWR 2005) – a project funded by the National Water Commission (NWC). Many individuals and organisations provided data for inclusion in the Water Account 2004-05 and the subsequent Experimental Estimates of Regional Water Use. The ABS wishes to acknowledge the contribution of Federal, state/territory and local government agencies; water authorities; and a range of private sector organisations and individuals that provided data for the accounts. Without their contribution the Water Account would not have been possible. Without the financial resources provided by the NWC the regional water use estimate project would not have been possible.
These documents will be presented in a new window.