Background
Statistical Surveys
States and Tribes are required by the Clean Water Act (CWA) to periodically report on the condition of all their waters. The States and Tribes typically monitor at targeted sites and can make scientifically-defensible statements about the condition of their waters only at these sites. Typically, this covers only a small percentage of the total waters. Ideally, an organization would survey an entire population, otherwise known as a census, to perfectly reflect the condition of a resource. However, this is rarely possible and organizations are forced to implement a Statistical (Probabilistic) Survey Design to randomly select a subset of the population to reduce the cost and effort when sampling. In such a design, each sampling site represents a specific portion of the total resource or population of interest such as all river and stream length in the nation. Because of the statistical nature of site selection, results from the sample population can be extrapolated to the entire population. For this reason, probability surveys are well suited for making unbiased assessments of the condition of an entire resource across large geographic areas with a known confidence.
NARS and Monitoring Initiative Funds
The Environmental Protection Agency (EPA) utilizes statistical survey designs in the National Aquatic Resource Surveys (NARS). NARS provides statistically valid national assessments of the condition of the nation’s waters (streams and rivers, lakes, coastal waters, and wetlands). EPA encourages states to implement a statistical survey in their monitoring program to provide comprehensive statements about statewide water quality conditions (in line with §305(b) reporting).
Monitoring Initiative funds are allocated to states, eligible interstate agencies, and eligible tribes using a formula that targets funds to enhance monitoring strategies and develop statistically valid reports on water condition. EPA, states and other partners collaborate to design the National Aquatic Resource Surveys, which cover the continental United States. For each sampling site selected through the survey design that falls within their jurisdiction, a state, eligible interstate agency, or eligible tribe receives approximately $8,000. Alaska, Hawaii, and the territories are not included in the National Aquatic Resource Surveys, but receive some of this funding for statistical survey projects.
The requirements for receiving monitoring initiative funds are:
1. Implementing a statistical survey design that provides condition estimates for a population of waters (e.g., streams, rivers, lakes, coastal waters, or wetlands) of the state based on an unbiased, representative sample of a subset of those waters. If they choose not to they must describe the program capacity development activities they are proposing and link them to their monitoring strategy.
2. Participate in the national/regional scale surveys (which may just be agreement to receive in-kind assistance).
3. Report the results of the states scale survey, preferably as a component of the state’s Integrated Report (narrative form) and/or in the probability survey module of ATTAINS.
Leveraging NARS for Partners
EPA encourages states, tribal nations, and other partners to leverage implementation of the NARS whether through the addition of indicators of special interest, sampling additional sites to achieve state-level or other program monitoring objectives or integrating their monitoring program with NARS.
EPA defines the addition of sites to the NARS as an intensification. Taking on an intensification typically requires significant coordination between the state, tribe or other groups involved and EPA to make the project a success. There are several things to think about when considering an intensification such as your overall objectives, costs, where the funding will come from and who will conduct the work. Below is more detail on these and other items. If you have decided that an intensification might be feasible, we encourage you to reach out to your Regional Monitoring Coordinator to talk through the next steps in planning and logistics.
Definitions
Auxiliary Variable - numeric attribute which represents the proportional (to size) inclusion probability variable.
Category - variables found within a stratum used to define design weights for unequal probability selections.
Census - examining every unit in a population of interest.
Generalized Random Tessellation Stratified (GRTS) Algorithm - is a spatially balanced sampling algorithm available in spsurvey.
Equal Probability Selection - selection where all units of the population have the same probability of being selected.
Non-Target Population - portion of the resource that is not of interest.
Population Estimate - estimates of the extent of a population, based on categorical and continuous variables.
Proportional Probability Selection - selection where the chance of being included is proportional to the values of a positive auxiliary variable.
Sample Frame - a GIS representation of the aquatic resource target population used in the selection of the sample.
Sample Population - the portion of the target population that might have been selected to be sampled and that could have been sampled in the field.
Stratum - a subpopulation within your sample frame to independently sample.
Spatial Balance - the degree of spatial spreading of a sample.
spsurvey - R package which provides functions for spatial survey design and analysis.
Statistical Survey - a cost effective, statistically representative method for assessing condition of a broad population and tracking changes over time.
Survey - a research method used for collecting data from a predefined group of sites.
Survey Weight - positive values associated with the observations in your dataset, used to ensure that metrics derived from a data set are representative of the population.
Target Population - an explicit definition of the resource of interest.
Unequal Probability Selection - selection where the chance of being included is set according to categorical variables.