Draft Project Deliverables: Policy Implications and Technical Basis

Size: px

Start display at page:

Download "Draft Project Deliverables: Policy Implications and Technical Basis"

Tiffany Webster
6 years ago
Views:

1 Surveillance and Monitoring Program (SAMP) Joe LeClaire, PhD Richard Meyerhoff, PhD Rick Chappell, PhD Hannah Erbele Don Schroeder, PE February 25, 2016 Draft Project Deliverables: Policy Implications and Technical Basis CV-SALTS Executive Committee Policy Meeting

2 Presentation Outline SAMP Objectives Direction from the August 2015 Executive Committee Meeting Well Selection Methodology Results Policy Issues Three policy questions for CV-SALTS Executive Committee to discuss. Detailed Technical Discussion to Follow 2

3 SAMP Overview The Recycled Water Policy also requires that the SNMPs include a Surveillance and Monitoring Program (SAMP) as a component: A basin/sub-basin wide monitoring plan that includes an appropriate network of monitoring locations. The scale of the basin/sub-basin monitoring plan is dependent upon the sitespecific conditions and shall be adequate to provide a reasonable, cost-effective means of determining whether the concentrations of salt, nutrients, and other constituents of concern as identified in the salt and nutrient plans are consistent with applicable water quality objectives. 3

4 SAMP Objectives Demonstrate the effectiveness of implementation of the SNMP through: Establishment of a cost-efficient program that relies on data from existing monitoring programs to the maximum extent possible. Use of selected data sources that provide a statisticallysignificant dataset to periodically assess: Ambient water quality (AWQ). TDS and nitrate trends in groundwater. 4

5 August 2015 Executive Committee Meeting Authorization to begin work: March 18, Discussed policy-related questions at the Executive Committee policy meeting on August 13, Outcomes included: SAMP will focus on groundwater basins underlying the Valley floor. The monitoring network should include wells other than community supply wells to avoid a bias towards better water quality. Ambient water quality and trends will be assessed every five years. 5

6 August 2015 Executive Committee Meeting Discussed policy-related questions at the Executive Committee policy meeting on August 13, Outcomes included: SAMP data will include groundwater elevation, TDS, nitrate, electrical conductivity and ancillary water quality data only. CV-SALTS database will be used as the basis for ambient water quality determinations. 6

7 Methodology for Well Network Selection Methodology outlined in the work plan and refined after August 2015 Executive Committee Meeting Well monitoring network based on IAZs. Use regular, equal-area grids to decluster data. Size of the grid a function of the data variability; larger grid cells can represent the data with fewer required wells. Accomplished with statistical tools. Assign land use to each grid. One well selected per grid cell (shallow and deep). Algorithm developed to select wells within each grid cell. Executed with geographic information system tools. 7

8 Power Analyses and Grid Cell Selection 8

9 Proposed SAMP: Deep Zone Apply well selection methodology, Goal is to minimize the difference between average water quality in SAMP network and the average using all wells. Increases confidence that data are representative. IAZ SAMP Wells - Average Nitrate All Wells - Average Nitrate % Diff SAMP Wells - Average TDS All Wells - Average TDS % Diff % % % % % % % % % % % % % % % % % % % 1,071 1,011 6% % % % % % % % 1,041 1,160-11% % % % % % % % % % % % % % % % 1,191 1,092 9% 9

10 Proposed SAMP: Deep Zone Goal is to minimize the difference between average water quality in SAMP network and the average using all wells. Increase confidence that data are representative. IAZ SAMP Wells - Average Nitrate All Wells - Average Nitrate % Diff SAMP Wells - Average TDS All Wells - Average TDS % Diff % % % % % % % % % % % % % % % % % % % 1,071 1,011 6% % % % % % % % 1,041 1,160-11% % % % % % % % % % % % % % % % 1,191 1,092 9% 10

11 Proposed SAMP: Deep Data Sources Selected Wells IAZ All Selected CDPH Popultion Served > 25 CDWR DAIRY GAMA USGS Number of Selected Wells Total Available % Total % % % % % % % % % % % % % % % % % % % % % % Total:

12 Proposed SAMP: Shallow one Goal is to minimize the difference between average water quality in SAMP network and the average using all wells. Increase confidence that data are representative. IAZ SAMP Wells - Average Nitrate All Wells - Average Nitrate % Diff SAMP Wells - Average TDS All Wells - Average TDS % Diff % % % % % % % 2,448 2,451 0% % % % 968 1,338-32% % 1, % % % % 1,123 1,301-15% % 1,096 1,182-8% % 941 1,249-28% % % % % % 4,849 4,461 8% % 3,045 1,505 68% % % % % % % % 9,569 9,589 0% % 1,043 1,026 2% % % % 1,422 1,322 7% 12

13 Proposed SAMP: Shallow Data Sources Selected Wells IAZ CDPH CDWR DAIRY GAMA USGS Number of Selected Wells Total Available % Total % % % % % % % % % % % % % % % % % % % % % % Total:

14 SAMP Deliverables Maps of SAMP monitoring network; deep and shallow Tables of proposed wells by IAZ Statistical and geographic information system tools 14

15 15

16 16

17 Policy Issue 1. Shallow Zone Data The well coverage in the shallow zone is not spatially extensive. There is a high degree of data variability. CV-SALTS will need to accept this variability or it may want to consider a single monitoring network for the production zone as an alternate approach. Provides assessment of ambient water quality where most water is extracted from. Reduction in monitoring costs. However, one would lose the resolution to assess water quality in shallow vs. deep. This modified approach is also relevant to Policy Issue 2. 17

18 Policy Issue 2. Implication of Limiting Data Bias Using only community water system wells biases the network towards better water quality; therefore the SAMP instead selected wells randomly when the overlying land use was not urban or semi-agricultural (Direction provided by CV-SALTS Executive Committee in August 2015). As a result: In the deep zone, between 67 and 80 percent of the wells are believed to be monitored routinely under existing programs. In the shallow zone, routine monitoring by others does not occur to the same extent as the deep zone Outcome: Avoiding bias in well selection increases the number of wells that will need to be annually sampled by this SAMP. This affects the cost of implementation. 18

19 Policy Issue 3. SAMP Start-up & Implementation Many questions regarding how the SAMP will be initiated. Examples of issue requiring resolution include: Need to ground-truth deep/shallow wells in recommended network, i.e., verify list of wells that will need to be sampled under this SAMP. Establish sampling and analysis protocols to ensure data consistency (accept other protocols or develop new ones). Determine how SAMP is to be funded. Basic costs include sample collection at selected wells not part of an existing monitoring network, laboratory analyses, data management (next slide), ambient water quality determinations, and periodic reporting. 19

20 Policy Issue 3. SAMP Start-up & Implementation Long-term implementation also requires resolution of a number of data management issues: Use of the existing CV-SALTS database or other databases. Need for database protocols for managing new data generated by the SAMP and providing the electronic data deliverables to the appropriate state databases. Development of queries to retrieve data for wells from the state databases that are already routinely sampled. Establishment of data QA/QC protocols for periodically evaluating uploaded data. Determining where database will be stored, e.g., Central Valley Water Board and at least two other entities. 20

21 Policy Issue 3. SAMP Start-up & Implementation Given the numerous issues that need to be addressed during SAMP start-up and long term implementation, it is recommended that CV-SALTS establish a project committee or work group during the basin plan amendment process to work through each of the SAMP implementation issues. 21

22 Surveillance and Monitoring Program (SAMP) Joe LeClaire, PhD Richard Meyerhoff, PhD Rick Chappell, PhD Hannah Erbele February 25, 2016 CV-SALTS Executive Committee Policy Meeting Technical Discussion

23 Presentation Outline Review TDS and nitrate spatial distribution Methods to decluster data Power analyses and grid cell selection IAZ-17 pilot area/algorithm for well selection for monitoring network Proposed SAMP for the Central Valley at the IAZ-Level 23

24 Presentation Outline Review TDS and nitrate spatial distribution Methods to decluster data Power analyses and grid cell selection IAZ-17 pilot area/algorithm for well selection for monitoring network Proposed SAMP for the Central Valley at the IAZ-Level 24

25 Scale of SAMP SAMP at the IAZ Level 20,533 total square miles Range is 282 to 1648 square miles with an average of 933 square miles. Local SNMP can define management zones with a MZlevel SAMP. 25

26 Data Distribution Shallow Zone Period of analysis: 2003 to wells with nitrate and TDS data 4567 wells with nitrate only 706 wells with TDS only 26

Data Distribution Deep Zone Period of analysis: 2003 to 2014 5310 wells with

27 Data Distribution Deep Zone Period of analysis: 2003 to wells with nitrate and TDS data 2693 wells with nitrate only 710 wells with TDS only 27

28 Data Distribution Unknown Zone Period of analysis: 2003 to wells with nitrate and TDS data 4 wells with nitrate only 131 wells with TDS only 28

29 TDS Areal Distribution Shallow Zone Period of analysis: 2003 to wells with TDS < 250 mg/l 561 wells with TDS between 250 and 500 mg/l 954 wells with TDS between 500 and 1000 mg/l 790 wells with TDS between 1000 and 5000 mg/l 60 wells with TDS between 5000 and 10,000 mg/l 45 wells with TDS > 10,000 mg/l 29

30 TDS Areal Distribution Deep Zone Period of analysis: 2003 to wells with TDS < 250 mg/l 2058 wells with TDS between 250 and 500 mg/l 765 wells with TDS between 500 and 1000 mg/l 245 wells with TDS between 1000 and 5000 mg/l 3 wells with TDS between 5000 and 10,000 mg/l 2 wells with TDS > 10,000 mg/l 30

31 Nitrate Areal Distribution Shallow Zone Period of analysis: 2003 to wells with nitrate < 2 mg/l 1054 wells with nitrate between 2 and 5 mg/l 988 wells with nitrate between 5 and 10 mg/l 927 wells with nitrate between 10 and 20 mg/l 624 wells with nitrate between 20 and 40 mg/l 376 wells with nitrate > 40 mg/l 31

32 Nitrate Areal Distribution Deep Zone Period of analysis: 2003 to wells with nitrate < 2 mg/l 2154 wells with nitrate between 2 and 5 mg/l 1306 wells with nitrate between 5 and 10 mg/l 592 wells with nitrate between 10 and 20 mg/l 286 wells with nitrate between 20 and 40 mg/l 89 wells with nitrate > 40 mg/l 32

33 Presentation Outline Review TDS and nitrate spatial distribution Methods to decluster data Power analyses and grid cell selection IAZ-17 pilot area/algorithm for well selection for monitoring network Proposed SAMP for the Central Valley at the IAZ-Level 33

34 Methods for Declustering Data Wells in the SAMP database are generally unequally distributed spatially across the IAZs Often contain both well clusters and areas without well coverage. It is therefore important that the method used to select wells for inclusion in the monitoring network incorporate the ability to simultaneously decluster and maximize coverage across the IAZ. 34

35 Methods for Declustering Data Three methods were considered: Equal-area and regularly-shaped (square in this case) grid cell method Equal-area and irregularly-shaped grid cell method developed by the USGS Unequal-area and irregularly-shaped grid cell method based on implementation of Thiessen polygons Second two methods can be spatially biased because the irregularly-shaped grid cells tending to estimate concentrations for large areas that lacked actual well coverage Equal-area and regularly-shaped was the selected method. 35

36 Presentation Outline Review TDS and nitrate spatial distribution Methods to decluster data Power analyses and grid cell selection IAZ-17 pilot area/algorithm for well selection for monitoring network Proposed SAMP for the Central Valley at the IAZ-Level 36

37 Power Analyses and Grid Cell Selection The equal-area, regularly-shaped grid cell method was selected because it focused on the actual data available and therefore minimized potential bias resulting from estimation across areas lacking well coverage. Square grid cells of various sizes were included for possible selection, ranging from 1 square mile (1 mile x 1 mile) to 16 square miles (4 mile x 4 mile). Selecting the most appropriate grid cell size for each IAZ was a function of the number of populated grid cells (number of grid cells containing at least one well) and the resulting variability in nitrate or TDS concentrations for the grid cell size and wells selected. 37

38 Power Analyses and Grid Cell Selection 38

39 Power Analyses and Grid Cell Selection 1. Select the grid cell size to evaluate, in sequence, beginning with the largest grid cell size (16 square mile grid) and ending with the smallest grid cell size (1 square mile grid). 2. Randomly select one well from each of the n populated grid cells. 3. Calculate the mean value of the n selected wells. 4. Repeat Steps 2-3 for m = 1000 bootstrap (with replacement) resamples. 39

40 Power Analyses and Grid Cell Selection 4. Calculate the mean of the resamples and determine the lower and upper confidence limits (LCL and UCL) of the mean as the 2.5th and 97.5th percentiles, respectively. 5. Calculate the lower and upper margins of error as the mean minus the LCL and the UCL minus the mean, respectively; and the lower and upper percent margins of error as the margins of error divided by the mean times Repeat Steps 1-6 for the next grid cell size, until all 10 grid cell sizes have been evaluated. 40

41 Power Analyses and Grid Cell Selection As the grid cell size decreases, the number of populated grid cells increases and the variability (margin of error) decreases. Therefore, the number of wells to include in the monitoring network depended on selecting a set of grid cell sizes that resulted in a practical and consistent margin of error across all IAZs (to the extent possible given the SAMP database). The target criterion for selection of a grid cell size was an upper margin of error (UME) of between 15 percent 41

42 Power Analysis for Mean of Avg NO3 (mg/l) : IAZ 17 : Deep Initial Analysis Zone (IAZ) : Depth Class : Deep Deep Deep Deep Deep Deep Deep Deep Deep Deep Grid Cell Size : 16_SQM_GRD 14_SQM_GRD 12_SQM_GRD 10_SQM_GRD 9_SQM_GRD 8_SQM_GRD 6_SQM_GRD 4_SQM_GRD 2_SQM_GRD 1_SQM_GRD Population Parameter : Mean Mean Mean Mean Mean Mean Mean Mean Mean Mean Confidence Level : Resample Size : Sample Size : Mean : Lower Confidence Limit : Upper Confidence Limit : Lower Margin of Error : Upper Margin of Error : % LME : % UME : Avg NO3 (mg/l) Upper Confidence Limit Mean Lower Confidence Limit 0 16_SQM_GRD 14_SQM_GRD 12_SQM_GRD 10_SQM_GRD 9_SQM_GRD 8_SQM_GRD 6_SQM_GRD 4_SQM_GRD 2_SQM_GRD 1_SQM_GRD 6 5 Margin of Error Upper (UCL - Mean) Lower (Mean - LCL) _SQM_GRD 14_SQM_GRD 12_SQM_GRD 10_SQM_GRD 9_SQM_GRD 8_SQM_GRD 6_SQM_GRD 4_SQM_GRD 2_SQM_GRD 1_SQM_GRD 42

43 IAZ Statistic Units 16 Square 14 Square 12 Square 10 Square 9 Square 8 Square 6 Square 4 Square 2 Square 1 Square Mile Grid Mile Grid Mile Grid Mile Grid Mile Grid Mile Grid Mile Grid Mile Grid Mile Grid Mile Grid 1 Mean TDS mg/l UME/Mean % Sample Size Mean TDS mg/l UME/Mean % Sample Size Mean TDS mg/l UME/Mean % Sample Size Mean TDS mg/l UME/Mean % Sample Size Mean TDS mg/l UME/Mean % Sample Size Mean TDS mg/l UME/Mean % Sample Size Mean TDS mg/l UME/Mean % Sample Size Mean TDS mg/l UME/Mean % Sample Size Mean TDS mg/l UME/Mean % Sample Size Mean TDS mg/l UME/Mean % Sample Size Mean TDS mg/l UME/Mean % Sample Size Mean TDS mg/l UME/Mean % Sample Size Mean TDS mg/l UME/Mean % Sample Size Mean TDS mg/l UME/Mean % Sample Size Mean TDS mg/l UME/Mean % Sample Size Mean TDS mg/l UME/Mean % Sample Size Mean TDS mg/l UME/Mean % Sample Size Mean TDS mg/l UME/Mean % Sample Size Mean TDS mg/l UME/Mean % Sample Size Mean TDS mg/l UME/Mean % Sample Size Mean TDS mg/l UME/Mean % Sample Size Mean TDS mg/l UME/Mean % Sample Size Square 14 Square 12 Square 10 Square IAZ Statistic Units Mile Grid Mile Grid Mile Grid Mile Grid 9 Square Mile Grid 8 Square Mile Grid 6 Square Mile Grid 4 Square Mile Grid 2 Square Mile Grid 1 Square Mile Grid 1 Mean TDS mg/l UME/Mean % Sample Size Mean TDS mg/l UME/Mean % Sample Size Mean TDS mg/l UME/Mean % Sample Size Mean TDS mg/l UME/Mean % Sample Size Mean TDS mg/l UME/Mean % Sample Size Mean TDS mg/l UME/Mean % Sample Size Mean TDS mg/l UME/Mean % Sample Size Mean TDS mg/l UME/Mean % Sample Size Mean TDS mg/l UME/Mean % Sample Size Mean TDS mg/l UME/Mean % Sample Size Mean TDS mg/l UME/Mean % Sample Size Mean TDS mg/l UME/Mean % Sample Size Mean TDS mg/l UME/Mean % Sample Size Mean TDS mg/l UME/Mean % Sample Size Mean TDS mg/l UME/Mean % Sample Size Mean TDS mg/l UME/Mean % Sample Size Mean TDS mg/l UME/Mean % Sample Size Mean TDS mg/l UME/Mean % Sample Size Mean TDS mg/l UME/Mean % Sample Size Mean TDS mg/l UME/Mean % Sample Size Mean TDS mg/l UME/Mean % Sample Size Mean TDS mg/l UME/Mean % Sample Size

44 Power Curve Analysis IAZ Statistic Units 16 Square 14 Square 12 Square 10 Square 9 Square 8 Square 6 Square 4 Square 2 Square 1 Square Mile Grid Mile Grid Mile Grid Mile Grid Mile Grid Mile Grid Mile Grid Mile Grid Mile Grid Mile Grid 1 Mean Nitrate mg/l UME/Mean % Sample Size Mean Nitrate mg/l UME/Mean % Sample Size Mean Nitrate mg/l UME/Mean % Sample Size Mean Nitrate mg/l UME/Mean % Sample Size Mean Nitrate mg/l UME/Mean % Sample Size Mean Nitrate mg/l UME/Mean % Sample Size

45 Power Curve Analysis IAZ Statistic Units 16 Square 14 Square 12 Square 10 Square 9 Square 8 Square 6 Square 4 Square 2 Square 1 Square Mile Grid Mile Grid Mile Grid Mile Grid Mile Grid Mile Grid Mile Grid Mile Grid Mile Grid Mile Grid 1 Mean Nitrate mg/l UME/Mean % Sample Size Mean Nitrate mg/l UME/Mean % Sample Size Mean Nitrate mg/l UME/Mean % Sample Size Mean Nitrate mg/l UME/Mean % Sample Size Mean Nitrate mg/l UME/Mean % Sample Size Mean Nitrate mg/l UME/Mean % Sample Size

46 Power Analyses and Grid Cell Selection First selected UME/Mean = 20 percent Revised to 15 percent Yields these grid cell sizes IAZ Deep Grids OLD Deep Grids NEW

47 Presentation Outline Review TDS and nitrate spatial distribution Methods to decluster data Power analyses and grid cell selection IAZ-17 pilot area/algorithm for well selection for monitoring network Proposed SAMP for the Central Valley at the IAZ-Level 47

48 Well Selection Algorithm One well per grid. Active status; if no active wells, then use inactive wells. Data between 2003 and 2014 Identify land use. Community land use CDPH/community water system wells given priority Semi-Agricultural: Dairy wells given priority Other land types: No well type are given priority selected by random number generator 48

49 Land use polygons Land use assigned to each grid cell 49

50 TDS distribution, using average of all wells in each grid cell. TDS distribution, using the selected well concentration in each grid cell. 50

51 Nitrate distribution, using average of all wells in each grid cell. Nitrate distribution, using the selected well concentration in each grid cell. 51

cell selection IAZ-17 pilot area/algorithm for well selection

52 Presentation Outline Review TDS and nitrate spatial distribution Methods to decluster data Power analyses and grid cell selection IAZ-17 pilot area/algorithm for well selection for monitoring network Proposed SAMP for the Central Valley at the IAZLevel 52

53 Proposed SAMP Steps for automation Assign land use Randomly select wells based on priority criteria Assign water quality value to grid cell Compute area weighted concentrations (for cells that are not entirely contained within an IAZ Estimate area weighted concentrations for SAMP wells Compare with concentrations using all of the wells. 53

54 Proposed SAMP ArcGIS model builder used to automate the well selection process 54

55 Proposed SAMP: Deep Zone Goal is to minimize the difference between average water quality in SAMP network and the average using all wells. Increase confidence that data are representative. IAZ SAMP Wells - Average Nitrate All Wells - Average Nitrate % Diff SAMP Wells - Average TDS All Wells - Average TDS % Diff % % % % % % % % % % % % % % % % % % % 1,071 1,011 6% % % % % % % % 1,041 1,160-11% % % % % % % % % % % % % % % % 1,191 1,092 9% 55

56 Proposed SAMP: Deep Data Sources Selected Wells IAZ All Selected CDPH Popultion Served > 25 CDWR DAIRY GAMA USGS Number of Selected Wells Total Available % Total % % % % % % % % % % % % % % % % % % % % % % Total:

57 Proposed SAMP: Shallow one Goal is to minimize the difference between average water quality in SAMP network and the average using all wells. Increase confidence that data are representative. IAZ SAMP Wells - Average Nitrate All Wells - Average Nitrate % Diff SAMP Wells - Average TDS All Wells - Average TDS % Diff % % % % % % % 2,448 2,451 0% % % % 968 1,338-32% % 1, % % % % 1,123 1,301-15% % 1,096 1,182-8% % 941 1,249-28% % % % % % 4,849 4,461 8% % 3,045 1,505 68% % % % % % % % 9,569 9,589 0% % 1,043 1,026 2% % % % 1,422 1,322 7% 57

58 Proposed SAMP: Shallow Data Sources Selected Wells IAZ CDPH CDWR DAIRY GAMA USGS Number of Selected Wells Total Available % Total % % % % % % % % % % % % % % % % % % % % % % Total:

59 Proposed SAMP Cumulative density function plot. Compares distribution of wells and average concentration in SAMP grid cells. Avg NO3 (mg/l) DepthClass = Deep : IAZ = Percent Selected_Value_Weighted Avg_Value_Weighted 59

60 Proposed SAMP Data clustering yields higher average ambient concentrations. Avg NO3 (mg/l) DepthClass = Deep : IAZ = Percent Selected_Value_Weighted Avg_Value_Weighted 60

61 Proposed SAMP Data clustering yields higher average ambient concentrations. SAMP 50% < 5 mg/l; all wells 50% < 7.5 mg/l SAMP: 75 percent of grids < MCL; all wells 60 percent of wells < MCL Avg NO3 (mg/l) DepthClass = Deep : IAZ = Percent Selected_Value_Weighted Avg_Value_Weighted 61

62 62

63 63

64 Questions? 64

Merger of the generator interconnection processes of Valley Electric and the ISO;

Merger of the generator interconnection processes of Valley Electric and the ISO; California Independent System Operator Corporation Memorandum To: ISO Board of Governors From: Karen Edson Vice President, Policy & Client Services Date: August 18, 2011 Re: Decision on Valley Electric