EPA Workshop: Probabilistic Monitoring

Size: px
Start display at page:

Download "EPA Workshop: Probabilistic Monitoring"

Transcription

1 EPA Workshop: Probabilistic Monitoring Virginia Engle Ecologist, U.S. EPA Presentation to Florida DEP Monitoring Program Q Meeting January 29, 2004

2 Aquatic Resources Monitoring Web Site

3 Clean Water Act Section 305(b) requires states to assess quality of all waters in biennial reports to EPA EPA summarizes findings in a national water quality inventory

4 Assessment Questions What is the ecological condition of waters in Florida? Has the proportion of coastal waters in poor condition changed since 1995? What is the extent of streams/rivers that meet their designated use? What proportion of lakes failed water quality standards in 2000?

5 Monitoring Designs Targeted or judgmental Census Statistical or probability-based survey

6 Targeted Monitoring Individual sites selected for a reason Cannot extrapolate to entire resource Cannot determine confidence levels Can determine condition of individual sites Examples State fixed-station networks USGS gaging stations Point-source discharge monitoring for NPDES

7 Census Condition measured at every single waterbody No error or uncertainty Assess site-specific condition and condition of all waters Impractical, expensive, time-consuming, resource intensive Examples US Population Census Remote sensing, GIS, mapping - census area & boundaries of entire resource

8 Probability-based Surveys Subset of waters randomly selected Can estimate condition of all waters Can calculate confidence levels Practical, cost effective, efficient, flexible Documented statistical principles Examples Opinion polls National Coastal Assessment Florida s IWRM and IMAP

9 OW Supports Probability-based Monitoring No State has sufficient monitoring resources to sample all its waters. With probability-based monitoring, a State can report assessment results for the target resource as a whole (e.g., all headwater streams) not just those waters that have been monitored. These assessment results are unbiased and include confidence limits. Source:

10 2002 Integrated Reporting Guidance A probabilistic monitoring design applied over large areas is an excellent approach to producing, with known confidence, a snapshot or statistical representation of the extent of waters that may or may not be impaired. States are encouraged to use probabilistic designs for water quality assessments and to include reports of these assessments in with their Integrated Reports. Source: Nov. 19, 2001 memo from Robert H. Wayland, Director, OWOW 2002 Integrated Water Quality Monitoring and Assessment Report Guidance

11 Survey Design Process Monitoring Objectives Institutional Constraints Survey Design Site Selection Design Requirements Target Population Sample Frame

12 Survey Design Types Simple random sample Systematic random sample GRTS Generalized Random Tessellation Stratified Spatially balanced Many options for nested subsamples, oversample, panels

13 Florida s Water Monitoring Objectives Characterize environmental quality of Florida s water resources IWRM provide scientifically defensible, statewide information on important chemical physical and biological characteristics of major surface water bodies and aquifer systems IMAP - assess the coastal marine waters of Florida using a set of key ecological indicators

14

15

16 Florida Watershed Basins and FDEP Districts

17 Florida Water Resources Resource Groundwater Aquifers Streams / Rivers Small Lakes Large Lakes Estuaries Design Type Point Linear Point Area Area

18

19 Groundwater Survey Design Reporting units = 5 WMD x 4 Subregions 5-year rotating basin with repeats Design Hex grid overlaid on subregion Random point selected in each hex List of well sites nearest random point Sampling location was chosen randomly from the list

20 Groundwater Aquifer Survey Design 2004 Objective: Estimate condition of confined and unconfined aquifers Estimate % wells failing chemical stds Estimate % area failing chemical stds Sample frame existing wells identified by basin and aquifer type Design GRTS spatially balanced for points stratified by basin and aquifer type

21 GRTS Design Sample size 30 wells per stratum Oversample 500% (150 wells) Inclusion probability = inverse of well density Weights Number of wells inverse IP Aquifer area actual if known or 100 units to estimate % aquifer area

22

23 Sarasota Peace-Myakka Basin

24 2004 Design

25 Groundwater Assessment < 2000 FL Ground Water Monitoring Network provided info on background water quality > 2000 GWQCI and BRI developed to assess human health impacts and aesthetic condition of groundwater on wellspecific and regional basis

26 2000 Groundwater Condition Confined Aquifers BRI(h) BRI(a) Percent "A" Wells Okeechobee Ocklawaha St. Marks Suwanee Tampa Source: Florida (b) Report

27 Rivers / Streams Design year rotating basin design ~50 random sites in 1 reporting unit per WMD each year Stratified by WMD, reporting unit, and stream order Redesign in 2004 Reporting units aligned with TMDL Management Rivers separate from streams Change to GRTS design

28 Rivers / Streams Design # Sites in 2000 WMD HS LS # Sites in 2001 WMD HS LS NW-A NW-B SF-A SF-D SJ-D SJ-B SR-A 52 2 SR-B SW-B SW-C HS=High order streams LS=Low order streams

29 Rivers / Streams 2000 Site Locations

30 Rivers and Streams Survey Design 2004 Objective: Assess condition of mgmt rivers and streams Estimate % km attaining designated use Sample frame Continuous stream network from RNHD 1:100K coverage Design Multi-Density GRTS for linear network stratified by rivers and streams Multi-Density Category reporting unit 30 sites per r.u. with 500% oversample

31 Florida Streams 2000 Northwest Florida WMD - A % Stream km Class III Criteria DO (mg/l) Screening Level

32 Florida Streams 2000 DO < 5.0 mg/l * % Stream km NW-A SF-A SJ-D SR-A SW-A * Doesn t meet criteria

33 Florida Streams 2000 DO < 5.0 mg/l WMD % meters meters Total m* NW-A 47.6% 1,025,784 2,155,719 SF-A 47.5% 1,971,493 4,149,086 SJ-D 39.8% 309, ,316 SR-A 55.0% 164, ,073 SW-B 61.3% 1,079,938 1,762,282 * Total meters assessed

34 Florida Streams 2001 DO < 5.0 mg/l * % Stream km NW-B SF-D SJ-B SR-B SW-C * Doesn t meet criteria

35 Florida Streams 2000 Fecal Coliform > 200/100 ml % Stream km NW-A SF-A SJ-D SR-A SW-B 4.3

36 Florida Streams 2001 Fecal Coliform > 200/100 ml % Stream km NW-B SF-D SJ-B SR-B SW-C

37 Florida Streams 2000 % Stream km NW-A SF-A SJ-D SR-A SW-B TSS (mg/l)

38 Florida Streams 2001 % Stream km NW-B SF-D SJ-B SR-B SW-C TSS (mg/l)

39 Florida 305(b) Rivers & Streams Fully Supporting Threatened Impaired Not Assessed Fully Supporting Threatened Impaired Not Assessed 2002 Cat 1 Cat 2 Cat 3a/3b Cat 3c/3d Cat 4/5 Not Assessed

40 Example - Indiana Full Support / Threatened Partial / Non Support Not Assessed Source -

41 Lakes Survey Design Reporting units = 5 WMD x 4 Subregions 5-year rotating basin with repeats Sample frame RF3 Small Lakes 1 to 10 hectares Large Lakes - > 10 hectares Design Small Lakes Random selection from list Large Lakes Grid with 1 random point per hexagon

42

43

44 Lakes Survey Design 2004 Objective: Assess condition of lakes Estimate % area attaining designated use Sample Frame RNHD 1:100K adjusted for wetlands Small Lakes 1 to <10 hectares Large Lakes - 10 hectares Design Small Lakes GRTS for points Large Lakes GRTS for areal resource

45 GRTS Design Sample size 30 sites per basin Oversample 500% (150 sites) Multi-density categories Small Lakes basin and lake size Large Lakes basin Weights Small Lakes number of lakes per basin Large Lakes area of lakes per basin

46 Estuaries Survey Design WMD, 19 sampling units based on NOAA EDA/CDAs 5-year rotating basin with repeats Sample Frame 477 estuaries delineated from USGS 1:100K DLGs and NOAA navigational charts Design Statewide 30 sites Intensive 30 sites in each of 5 sampling units per year.

47 Estuaries Sampling Units and Schedule WMD NW Apalachicola Bay St. Andrew Bay Choctawhatchee Bay Pensacola Bay SJR Nassau/ St. Johns/ St. Marys Halifax/ Mosquito Lagoon IRL North/ Banana River IRL South SR Suwannee Sound Wacassassa Bay Big Bend Suwannee Sound SW Tampa Bay Sarasota/ Lemon Bays Charlotte Harbor West Central Coast S Lake Worth North 10K Islands South 10K Islands Florida/ Biscayne Bays

48 Florida Statewide Design

49

50

51 Florida 2000 Total Benthic Abundance Florida Base Sites Apalachicola Bay Percent Area Percent Area Benthic Abundance Benthic Abundance

52 Florida 2000 Total Benthic Abundance Lake Worth Nassau/St. Johns Percent Area Percent Area Benthic Abundance Benthic Abundance

53 Florida 2000 Total Benthic Abundance Suwanee Sound Tampa Bay Percent Area Percent Area Benthic Abundance Benthic Abundance

54 Florida 2000 Total Benthic Abundance State & Region Comparisons STATE TAM SUW NSJ LKW < >200 APA 0% 20% 40% 60% 80% 100% Percent Area

55 Florida 2000 Total Benthic Taxa Richness State & Region Comparisons STATE TAM SUW NSJ LKW < >25 APA 0% 20% 40% 60% 80% 100% Percent Area

56 Sediment Contaminants Indicator # Sites % Area # Sites % Area Hg > PQL 3 13 ± ± 11 Hg > TEL 2 3 ± ± 9 Metals > PQL ± ± 18 Metals > TEL 4 17 ± ± 12 Phenols > PQL 1 3 ± ± 7 Source: Florida IMAP Annual Report: Year 4

57 Florida 305(b) Estuaries Fully Supporting Threatened Impaired Not Assessed Fully Supporting Threatened Impaired Not Assessed 2002 Cat 1 Cat 2 Cat 3a/3b Cat 3c/3d Cat 4/5 Not Assessed