Maryland Department of Environment v. Anacostia Riverkeeper

ADKINS, J. FACTS AND LEGAL PROCEEDINGS Maryland Department of the Environment (“MDE”) issued municipal separate storm sewer system (“MS4”) discharge permits (“the Permits”) to Anne Arundel County, Baltimore City, Baltimore County, Montgomery County, and Prince George’s County (“the Counties”). Multiple organizations argue that the Permits do not comply with federal and state law, and they request that we remand for MDE to correct these legal errors. 1 ⁹⁶ Federal Framework: NPDES Permits and Municipal Stormwater Discharges Under the Clean Water Act (“CWA”), the discharge of pollutants is illegal. 33 U.S.C. § 1311 . Through the National Pollution Discharge Elimination System (“NPDES”), 2 33 U.S.C. § 1342 , either the Environmental Protection Agency (“EPA”) or an EPA-approved state, such as Maryland, may issue permits exempting a discharger from this prohibition. See Piney Run Pres.

Ass’n v. Cnty. Comm’rs of Carroll Cnty., Md., 268 F.3d 255, 265 (4th Cir.2001). MDE is the authority in Maryland that administers the NPDES program. Code of Maryland Regulations (“COMAR”) 26.08.04.07.

An NPDES permit, however, does not give a discharger carte blanche. “Generally speaking, the NPDES requires dischargers to obtain permits that place limits on the type and quantity of pollutants that can be released into the Nation’s waters.” S. Fla. Water Mgmt. Dist. v. Miccosukee Tribe, 541 U.S. 95, 102 , 124 S.Ct. 1537 , 158 L.Ed.2d 264 (2004). These limits are called effluent limitations.

See 33 U.S.C. § 1362 (11) (defining an effluent limitation as “any restriction established by a State or the Administrator on quantities, rates, and concentrations of chemical, physical, biological, and other constituents which are discharged from point sources into navigable waters, the waters of the contiguous zone, or the ocean, including schedules of compliance”). The type of discharge determines the type of limitations the permit must impose on the discharger. The Permits before us control stormwater pollutant dis ⁹⁷ charge. 3 Stormwater consists of the rain and snowmelt that filters through the soil and courses over surfaces — collecting pollutants along the way — before passing through the municipal storm sewer systems 4 and into waterbodies. During the development of the Permits, the Water Groups explained the problems that stormwater poses, whether to the surface conditions of Maryland’s waters, for humans who recreate and subsist on them, and for wildlife who live in them.

See Letter with Comments on Draft MS4 Permit for Baltimore City from Blue Water Baltimore, Inc. and Earthjustice, to Brian Clevenger, MDE (Sept. 21, 2012). In recognition of extensive public commentary on the severity of the problems associated with stormwater, MDE stated: “[i]t becomes fairly easy for all organizations, individuals, and government agencies to agree that urban stormwater is a problem that must be addressed.” MDE, Response to Formal Comments for Montgomery County NPDES Permit (2009). Nevertheless, municipal stormwater discharge is “highly intermittent,” “usually characterized by very high flows occurring over relatively short time intervals,” and “depend[s] on the activities occurring on the lands.” See National Pollutant ⁹⁸ Discharge Elimination System Permit Application Regulations for Storm Water Discharges, 55 Fed.Reg. 47,990, 48,038 (Nov. 16, 1990) (codified at 40 C.F.R. § 122.26 ). It is also difficult to discern the amount of pollutant that any one discharger contributes to a waterbody because municipalities have so many outfalls, or discharge points, leading into the waters.

See MDE, Montgomery County NPDES Permit Fact Sheet (900 outfalls); MDE, Anne Arundel County NPDES Permit Fact Sheet (nearly 1,000 outfalls); MDE, Baltimore County NPDES Permit Fact Sheet (nearly 700 outfalls.); MDE, Prince George’s County NPDES Permit Fact Sheet (more than 4,000 outfalls); MDE, Baltimore City NPDES Permit Fact Sheet (around 350 outfalls.); see also 40 C.F.R. § 122.26 (b)(5), (9) (outlining minimum diameters of pipes in major MS4 outfalls). 5 Because of the nature of municipal stormwater discharges, Congress adopted a flexible approach to the control of pollutants in MS4s. See 55 Fed.Reg. at 48,038 (The Congressional Record from 1986 stated not only that “an end-of-the-pipe treatment technology is not appropriate for [the MS4] discharge” but also that “[MS4] controls may be different in different permits.”). 6 Pursuant to 33 U.S.C. ⁹⁹ § 1342(p)(3)(B)(iii), municipal stormwater permits “shall require controls to reduce the discharge of pollutants to the maximum extent practicable, including management practices, control techniques and system, design and engineering methods ....” 7 Best management practices (“BMPs”) have been a longstanding control or effluent limitation 8 in MS4 permits. See 40 C.F.R. § 122.44 (k)(2) (BMPs “control or abate the discharge of pollutants when [a]uthorized under [ 33 U.S.C. § 1342 (p) ]”); id. § 122.44(k)(3) (BMPs are an appropriate control when “[n]umeric effluent limitations are infeasible”); see also Tualatin Riverkeepers v. Or. Dep’t of Envtl.

Quality, 235 Or.App. 132 , 230 P.3d 559, 564 (2010) (“ ‘Best management practices,’ such as those incorporated in the permits at issue in this case, are a type of effluent limitation.”). The EPA defined BMPs to mean “schedules of activities, prohibitions of practices, maintenance procedures, and other management practices to prevent or reduce the pollution of ‘waters of the United States.’ ” 40 C.F.R. § 122.2 ; cf. Natural Res. Def. Council, Inc. v. EPA, 808 F.3d 556, 579 (2d Cir.2015) (“But EPA’s narrative WQBEL [water-quality based effluent limitation] does not qualify as a BMP, as it is neither a practice nor a procedure.”).

Examples of the types of BMPs the Counties might implement pursuant to the Permits are infiltration practices and green roofs. 9 ¹⁰⁰ Through guidance memos, the EPA has endorsed the use of BMPs in MS4s for decades but has increasingly recommended that, where feasible, such permits include numeric effluent limitations. Interim Permitting Approach for Water-Quality Based Effluent Limitations in Storm Water Permits, 61 Fed. Reg. 48,761 (1996); EPA, Memorandum on Establishing Total Maximum Daily Load (TMDL) Wasteload Allocations (WLAs) for Storm Water Sources and NPDES Permit Requirements Based on Those WLAs § 3 (2002) [hereinafter “2002 Memo”]; EPA, Memorandum on Revisions to the November 22, 2002 Memorandum 4-5 (2010) [hereinafter “2010 Memo”]; EPA, Memorandum on Revisions to the November 22, 2002 Memorandum at 4 n. 5 (2014) [hereinafter “2014 Memo”]. (A “numeric” effluent limitation “refers to [a] limitation[] with a quantifiable or measurable parameter related to a pollutant (or pollutants).”). Total Maximum Daily Loads (“TMDLs”) The concept of total maximum daily load (“TMDL”) looms large in this case. 10 We begin by setting forth its basic purpose, then unpacking its complex formation.

TMDLs inform. See Am. Farm Bureau Fed’n v. EPA 792 F.3d 281, 291 (3d Cir.2015) (“Our understanding of [TMDLs] as informational tools is supported by every case and piece of scholarship to consider them as well as the language of the Chesapeake Bay TMDL itself.”); see also EPA, Chesapeake Bay TMDL § 1.41, at 1-15 (2010) (“TMDLs are ‘primarily informational tools’ that ‘serve as a link in an implementation chain____’ ”), available at http://www.epa.gov/chesapeake-bay ¹⁰¹ tmdl/ehesapeake-bay-tmdl-document [https://perma.cc/9R7V-9 VHV]. TMDLs arise out of a multi-step process that begins with the establishment of water quality standards (“WQS”).

See Am. Farm Bureau Fed’n, 792 F.3d at 289 (“TMDLs happen after a state enacts pursuant to its law (but required by the Clean Water Act) ‘water quality standards.’ ”). Because the EPA and the states interact throughout this process, it has been described as one of “cooperative federalism.” Id. at 288 ; see Anacostia Riverkeeper, Inc. v. Jackson, 798 F.Supp.2d 210, 214-17 (D.D.C.2011). Water quality standards, as the term itself suggests, protect water quality. 40 C.F.R. § 130.2 (d); COMAR 26.08.02.01(A).

Each state must set water quality standards by assigning a “use” to a water, such as recreation or fishing, then developing criteria to protect those uses, as well as ensuring that higher quality waters do not degrade to the minimally accepted standard (also known as an anti-degradation policy). 33 U.S.C. § 1313 ; COMAR 26.08.02.01(B)(1). All water quality standards are subject to EPA review, and if the EPA does not approve of them, the EPA will set those standards itself. 33 U.S.C. § 1313 . By way of example, the EPA approved a TMDL that MDE submitted for fecal bacteria for the Non-tidal Cabin John Creek Basin in Montgomery County in 2007. See MDE, Total Maximum Daily Loads of Fecal Bacteria for the Non-tidal Cabin John Creek Basin in Montgomery County, Maryland (Document version: Oct. 13, 2006) [hereinafter “John Creek Basin TMDL”].

These bacteria are microscopic organisms in animal waste. Fecal bacteria in water can raise the risk of illness in humans who recreate there. Id. at § 1.0, at 1. To develop the WQS, MDE selected “water contact recreation and protection of aquatic life and public water supply” as the use of the water and 126 MPN 11 per 100 milliliters as the criteria.

Id. at § 2.3, at 11 (citing COMAR 26.08.02.08(0) and ¹⁰² 26.08.02.03-3). This figure (126) represents a mean density for this pollutant. Id.; see 40 C.F.R. § 130.2 (i) (“[M]ass per time, toxicity, or [an]other appropriate measure” may be used to express TMDLs.). After setting WQSs, the states establish effluent limitations in permits as the primary way to meet the WQSs because, as we have explained, effluent limitations restrict the discharge of pollutants.

See 33 U.S.C. § 1362 (11). Nevertheless, we note, importantly, that MSbs are not subject to the requirement of imposing effluent limitations “necessary to meet water quality standards.” See 33 U.S.C. § 1311 (b)(1)(C); see also Defenders of Wildlife v. Browner, 191 F.3d 1159, 1165 (9th Cir.1999); cf. 33 U.S.C. § 1342 (p)(3)(A) (Industrial dischargers must comply with 33 U.S.C. § 1311 .). 12 This important point notwithstanding, Congress requires that “[e]ach State shall identify those waters within its boundaries for which the effluent limitations required by [ 33 U.S.C. § 1311 ] are not stringent enough to implement any water quality standard applicable to such waters.” 33 U.S.C. § 1313 (d)(1)(A). This is where the TMDL comes into play. The TMDL tells a state what is the threshold amount of a pollutant that a body of water can tolerate before violating the WQS.

See In re City of Moscow, Idaho, 10 E.A.D. 135 , 2001 WL 988721 , at ⁴ (EAB July 27, 2001) (“A TMDL is a measure of the total amount of a pollutant from point sources, nonpoint sources and natural background, that a water quality limited segment can tolerate without violating the applicable water quality standards.”); EPA, Chesapeake Bay TMDL § 1.1, at 1-2 (“A TMDL specifies the maximum amount of a pollutant that a waterbody can receive and still meet applicable WQS.”). 13 ¹⁰³ States must establish TMDLs “at a level necessary to implement the applicable water quality standards,” 33 U.S.C. § 1313 (d)(1)(C), when they identify those waters for which effluent limitations cannot implement the WQSs, 33 U.S.C. § 1313 (d)(1)(A). 14 As with water quality standards, the states have the obligation of setting TMDLs and submitting them to the EPA for approval. See supra MDE, John Creek Basin TMDL (The EPA approved of MDE’s TMDL in March 2007.). If the EPA disapproves of the TMDLs, the EPA will set them itself. 33 U.S.C. § 1313 (d)(2). For this case, wasteload allocations (“WLAs”) are the most critical part of the TMDL equation.

See 40 C.F.R. § 130.2 (i) (A TMDL is “[t]he sum of the individual WLAs for point sources and LAs [load allocations] for nonpoint sources and natural background.”). The WLA represents a water’s “loading capacity” assigned to its “point sources of pollution.” Id. § 130.2(h). Continuing with our example, MDE set the TMDL for fecal bacteria at 176.36 billion MPN/day, the LA at 68.17 billion MPN/day, and the WLA at 108.19 billion MPN/ day. MDE, John Creek Basin TMDL. 15 ¹⁰⁴ Although TMDLs are informational tools, of which WLAs are a part, WLAs are more akin to restrictions.

See Am. Farm Bureau Fed’n v. EPA 984 F.Supp.2d 289, 328 (M.D.Pa. 2013) (“WLAs are not permit limits per se; rather they still require translation into permit limits ....”) (citation omitted) (internal quotation marks omitted) (emphasis in original), aff'd, 792 F.3d 281 (3d Cir.2015). Under 40 C.F.R. § 122.44 (d)(1)(vii)(B), permitting authorities must ensure that effluent limitations “are consistent with the assumptions and requirements” of any approved WLA. 16 We conclude our introduction of TMDLs by noting that MS4s are subject to the MEP standard under 33 U.S.C. § 1342 . MS4s are not, however, required to impose effluent limitations necessary to meet water quality standards.

The CWA still requires Maryland to set water quality standards and TMDLs — subject to the EPA’s approval. Flowing from this obligation is the requirement that MS4s are subject to effluent limitations that are consistent with WLAs of EPA-approved TMDLs. The Chesapeake Bay TMDL As we will discuss in more detail, the Permits require the Counties to take actions to make progress in meeting the WLAs of many EPA-approved TMDLs. 17 By far, though, the ¹⁰⁵ most critical TMDL in this case is the Chesapeake Bay TMDL ("Bay TMDL”). Regarded as a national treasure, 18 the Chesapeake Bay is the largest estuary in the United States, a product of flooding from the Susquehanna River over thousands of years.

Alice Jane Lippson, The Chesapeake Bay in Maryland: An Atlas of Natural Resources 2 (Johns Hopkins University Press 1973). Over 2,000 species of animals and plants reside in the Chesapeake Bay. Alice Jane Lippson & Robert L. Lippson, Life in the Chesapeake Bay viii (Johns Hopkins University Press 1984). These include phytoplankton, the blue crab, and striped bass, among many, many others.

Lippson, The Chesapeake Bay in Maryland, supra at 14, 26, 36. In addition to housing much wildlife, the Chesapeake Bay is a shipping and commerce hub and a source of recreation. Chesapeake Bay Program, Chesapeake Bay: Introduction to an Ecosystem 2 (2004). Human activity, however, threatens this complex ecosystem. “Excess sediment and nutrients endanger the Bay’s water quality.” Id. at 3-4.

Such threats include: depriving species of oxygen; delivering chemicals which collect in animal tissue; and even destroying habitats because sunlight cannot reach critical underwater grasses where species reside. Id. There is, then, no underestimating the importance of the restoration of the Chesapeake Bay in Maryland. See Am.

Farm Bureau, 984 F.Supp.2d at 298 (“[The Bay] has been ¹⁰⁶ described as one of the most biologically productive ecosystems in the world,” and, along with its watersheds, “add[s] ecological, economic, recreational, historic, and cultural value to the region.”). How to restore the Bay, however, has been a prolonged, frustrated process. See id. (The Bay TMDL “is not a new or recent idea,” and thus, “it would be improper to view the Final TMDL in a vacuum as a single, isolated effort to restore water quality to the Chesapeake Bay.”).

Some of these restoration efforts include the Chesapeake Bay Agreement in 1980, another agreement in 1987, amendments to the agreement in 1992, and the Chesapeake 2000 Agreement. Department of Legislative Services, Office of Policy Analysis, Chesapeake Bay Restoration and the Tributary Strategy: An Analysis of Maryland’s Efforts to Meet the Nutrient and Sediment Reduction Goals of the Chesapeake 2000 Agreement 3^t (2007). The EPA established the Bay TMDL in December 2010. See Am.

Farm Bureau Fed’n, 792 F.3d at 290 (“As noted, for the Chesapeake Bay the relevant states and the EPA agreed that the EPA would draft the TMDL in the first instance.”). It has survived legal challenges before the U.S. District Court for the Middle District of Pennsylvania as well as the Third Circuit. 19 See Am. Farm Bureau Fed’n, 984 F.Supp.2d at 294 ; Am. Farm Bureau Fed’n, 792 F.3d at 287 .

These courts have noted that the efforts to restore the Chesapeake Bay extend back decades, and that the development of the Bay TMDL itself has been a decade-long process. Am. Farm Bureau Fed’n, 984 F.Supp.2d at 299 ; Am. Farm Bureau Fed’n, 792 F.3d at 291 .

The Bay TMDL provides information pertaining to pollution reduction for nitrogen, phosphorus and sediment in the Chesa ¹⁰⁷ peake Bay and applies to the District of Columbia and the six “Bay” states, including Maryland. EPA, Chesapeake Bay TMDL at ES-1. 20 Modeling Before delving into Maryland’s role in the formation of the Bay TMDL, we must discuss the “critical and valuable” role that modeling played in the Bay TMDL’s development. EPA, Chesapeake Bay TMDL at ES-5. 21 “Modeling is an approach that uses observed and simulated data to replicate what is occurring in the environment to make future predictions.” Id. “A model ‘is an abstraction from and simplification of the real world.’ ” Am. Farm Bureau, 984 F.Supp.2d at 340 (citation omitted).

Models are essential when one seeks to study “ecosystems that are too large or complex for real-world monitoring,” such as the Chesapeake Bay and its watersheds. Chesapeake Bay Program, About the Bay Program: Modeling, available at http://www.chesapeakebay.net/about/ programs/modeling [https://perma.ee/RDM4-C4XF] (last visited Feb. 9,2016) [hereinafter CBP: Modeling]. A prominent component in the modeling of the Bay TMDL was the Phase 5.3 Chesapeake Bay Watershed Model [“Phase 5.3 Model”]. EPA, Chesapeake Bay TMDL, at 5-19. 22 “The ¹⁰⁸ Phase 5.3 Model is the most recent of a series of increasingly-refined versions of the Chesapeake Bay Watershed Model.” EPA, Chesapeake Bay Phase 5.3 Community Watershed Model § 1.2.1, at 1-13 (2010), available at http://www.chesapeake bay.net/about/programs/modeling/53/ [https://perma.cc/9YKSWBZ4]. 23 The Phase 5.3 Model simulates the “loading and transport of nitrogen, phosphorus, and sediment from pollutant sources throughout the Bay watershed.” EPA, Chesapeake Bay TMDL, at 5-20.

Additionally, this model provides “estimates of watershed nitrogen, phosphorus, and sediment loads resulting from various management scenarios.” Id. Because models are not “perfect forecasts,” however, modeling is “part of a broader toolkit,” including monitoring, “to gain the highest possible level of accuracy.” CBP: Modeling. As the EPA explained: “The Bay modeling framework takes advantage of decades of atmospheric deposition, streamflow, precipitation, water quality, biological resource, and land cover monitoring data” as well as “tracking and reporting of the implementation of pollution load reduction best management practices.” EPA, Chesapeake Bay TMDL, at § 5.1, 5-1-5-2. These resources allowed the EPA to calibrate its models.

Id. Because the Bay TMDL exists in significant part as a result of modeling, and because of how prevalent modeling is in TMDL formulation, MDE incorporated modeling into the Permits. MDE incorporated by reference a document the agency published, called Accounting for Stormwater Wasteload Allocations and Impervious Acres Treated, Guidance for National Pollutant Discharge Elimination System Stormwater Permits ¹⁰⁹ (“the Guidance”). 24 As the name suggests, the document serves dual purposes: the Counties can assess progress in achieving WLAs and also assess restoration of impervious surface areas through a credits-to-acres approach. 25 In the Guidance, MDE sets forth acceptable models that the Counties can use, including, Maryland’s Assessment and Scenario Tool (“MAST”) 26 MDE, Guidance at 2. The Guidance includes the pollutant rates for the Bay TMDLs — Total Nitrogen (“TN”), Total Phosphorus (“TP”), and Total Suspended Sediment (“TSS”) — and requires that the Counties use these pollutant rates together with land use data to calculate baseline stormwater loads.

Id. at 2-3. As the document explains, “[t]hese pollutant loads are specific to the [ ] Bay TMDL.” Id. at 2. But the Counties may use the principles and methods in the Guidance “for any EPA approved TMDL.” Id. at 1. Maryland’s Watershed Implementation Plan (“WIP”) The EPA developed the Bay TMDL to ensure that the Bay jurisdictions would put in place “all pollution control measures needed to fully restore the Bay and its tidal rivers” by 2025.

EPA, Chesapeake Bay TMDL at ES-1. The EPA approved the Bay TMDL “only after” determining that each jurisdiction provided “reasonable assurance” that it would meet established pollutant reductions. Am. Farm Bureau Fed’n, 792 F.3d at 291 .

The Bay jurisdictions set forth their strategies for meeting pollutant reductions in Watershed Implementation Plans (“WIPs”). Id. WIPs are “roadmaps” setting forth a plan for how and when a jurisdiction will reach the pollution reduction goals in the Bay TMDL. EPA, Chesapeake Bay TMDL, at ES-8.

The EPA described these roadmaps as the “cornerstone” that ¹¹⁰ ensured the States were accountable in achieving pollution reductions. Id. Notably, the EPA expressed no concerns about Maryland’s Final WIP, 27 whereas the EPA had to implement backstop allocations and adjustments in other Bay jurisdictions so that the EPA had reasonable assurance that all jurisdictions would achieve necessary reductions. See EPA, Chesapeake Bay TMDL Executive Summary, ES-10-ES-13 (Dec. 29, 2010) [hereinafter “Bay TMDL ES”]. 28 Maryland’s WIP lists restoration of “twenty percent of the counties’ impervious surface area that is not already restored to the maximum extent practicable (MEP)” in the “key elements” supporting the reasonable assurance of the implementation of the WIP.

Phase I WIP at 5-80. The elements also include the adaptive management approach whereby additional or alternative practices are implemented if existing programs are not meeting target reductions. Id. 29 Maryland’s Stormwater Management History In addition to an explication of the federal permitting system, NPDES, and the complex components arising out of it, such as TMDLs, we also set forth Maryland’s stormwater ¹¹¹ management program, which has evolved since its inception in the 1980s, and which is informative for purposes of analyzing the Permits. In 1982, the General Assembly enacted laws “to reduce as nearly as possible the adverse effects of stormwater runoff.” Maryland Code (1982, 2007 Repl.Vol.), § 4-201 of the Environment Article (“EN”); see H.B. 1091, 1982 Gen.

Assemb. Reg. Sess. (Md.1982).

As a result, each county and municipality in Maryland was required for the first time to “adopt ordinances necessary to implement a stormwater management program” by July 1, 1984. See EN § 4-202. Then authorized by the General Assembly, the Department of Natural Resources issued regulations setting forth minimum control requirements and design criteria for the counties and municipalities. See 10 Md. Reg. 881 , 884-85 (May 13, 1983) (to be codified at COMAR 08.05.05). 30 The regulations fostered the “primary goal” of “maintaining] after development, as nearly as possible, the predevelopment runoff characteristics” of the land.

See EN § 4 — 203(b)(1); see also EN § 4-204(a) (Development of land is prohibited without submitting a stormwater management plan and obtaining the municipality’s or county’s approval of the plan.) Maryland entered a new phase of stormwater management in the early 2000s. Pursuant to EN § 4-203(b), MDE adopted regulations to “rectify the[] programmatic shortcomings” of then-existing regulations that had provided “sparse guidance” on “water quality enhancement.” 27 Md. Reg. 1167 , 1168 (June 16, 2000) (to be codified at COMAR 26.07.02). Amending the stormwater regulations, MDE intended to “provide water quality treatment of up to 90 percent of the average annual rainfall throughout the State, establish ground water recharge standards, and outline a channel erosion control strategy,” as well as “promote environmentally friendly site design.” Id. To fulfill this purpose, MDE incorporated by ¹¹² reference the 2000 Maryland Stormwater Design Manual (“the Manual”).

Id. at 1167, 1169 . 31 MDE required the counties and municipalities to revise their ordinances to incorporate the Manual’s policies and practices by July 1, 2001. Id. at 1170 . The Manual “provide[d] designers a general overview on how to size, design, select and locate BMPs at a new development site to comply with State stormwater performance standards.” Center for Watershed Protection (“CWP”) & MDE, Manual, § 1.3, at 1.16. There are 14 performance standards, including the water quality volume standard (“WQ,”).

Id. § 1.2, at 1.13. Another stormwater management phase began when the General Assembly required MDE to mandate the use of environmental site design (“ESD”) in 2007. H.B. 786, Gen. Assemb.

Reg. Sess. (Md.2007). ESD is best understood as those practices, such as “small-scale stormwater management practices, nonstructural techniques, and better site planning,” that “mimic natural hydrologic runoff characteristics and minimize the impact of land development on water resources.” EN § 4-201.1(b); see, e.g., note 9 (green roofs).

MDE implemented regulations to this effect and explained that “[t]he goal of the regulations is to maintain after development as nearly as possible, the predevelopment runoff characteristics of the site being developed using ESD to the MEP.” 35 Md. Reg. 2191 (Dec. 5, 2008) (to be codified at COMAR 26.17.02). The Permits MDE issued several series of MS4 permits to the Counties that preceded the Permits before us today. See MDE, NPDES MS4 Permit Montgomery County Fact Sheet (2008) ¹¹³ (The first two permits were issued in 1996 and 2001.); MDE, Basis for Final Determination to Issue Prince George’s County’s NPDES MS4 Permit (2013) (The first three permits were issued in 1993, 1999, and 2004.); MDE, Basis for Final Determination to Issue Baltimore County’s NPDES MS4 Permit (2013) (The first three permits were issued in 1994, 2000, and 2005.); MDE, Basis for Final Determination to Issue Anne Arundel County’s NPDES MS4 Permit (2013) (The first three permits were issued in 1993, 1999, and 2004.); MDE, Basis for Final Determination to Issue Baltimore City’s NPDES MS4 Permit (2013) (The first three permits were issued in 1993, 1999, and 2005.). 32 At issue here are five-year term Permits MDE most recently issued: to Montgomery County in February 2010, to Baltimore County in December 2014, to Baltimore City in December 2013, to Prince George’s County in January 2014, and to Anne Arundel County in February 2014. The Water Groups challenge the Permits in several respects, namely, (1) the requirement to restore impervious surface area, (2) the requirement to submit plans for TMDLs, (3) the monitoring requirements, and (4) the public’s ability to participate in the development of the Permits.

Some of these provisions are new and therefore represent an increase in responsibility on the Counties to maintain and improve the quality of their waters. See, e.g., MDE, Basis for Final Determination to Issue Anne Arundel County’s NPDES MS4 Permit (2013) (“These meetings resulted in the addition of more stringent conditions to Anne Arundel County’s storm-water permit, in large part due to a regional and growing focus on restoring Chesapeake Bay.”); id. (“New requirements in the permit will include ... developing restoration plans to meet stormwater WLAs for impaired waters.”); MDE, NPDES Montgomery County Stormwater Permit Response to Formal Comments at 2 (2009) (“MDE believes that ¹¹⁴ this current municipal stormwater permit will force Montgomery County to make major strides toward controlling urban runoff better than ever before. New conditions such as trash abatement jurisdiction-wide and requiring an additional 20% of the County’s impervious area to be restored are major additions.”); 33 MDE, Maryland’s 2006 TMDL Implementation Guidance for Local Governments i (2006) (“Until recently, Maryland has focused primarily on TMDL development, which establishes limits on pollutant loads.

Now the State is moving into the implementation phase....”). Before discussing these Permit provisions, we note additional Permit requirements that illustrate the breadth of the Counties’ obligations. The Counties must implement management programs “to control stormwater discharges to the maximum extent practicable.” These programs include a stormwater management program (“SWMP”) and an erosion and sediment control program in accordance with state law; an “illicit discharge detection and elimination” program; requirements to reduce trash; obligations on the Counties to reduce pollutants associated with maintenance activities and on municipal facilities to submit pollution prevention plans; as well as a requirement to engage in public outreach activities to reduce stormwater pollution. The Permits also require the Counties to engage in thorough analyses of the water quality of their watersheds.

Among other things, the watershed assessments oblige the Counties to identify and prioritize water quality improvement projects. MDE has also ensured that the Counties cannot use lack of adequate funding as a defense for failure to comply with Permit terms. The Permits explain that “[l]ack of funding ¹¹⁵ does not constitute a justification for noncompliance with the terms of this permit.” 34 To this end, the General Assembly enacted EN § 4-202.1 in 2012, requiring the Counties “to adopt local laws or ordinances necessary to establish an annual stormwater remediation fee and a local watershed protection and restoration fund to provide financial assistance for the implementation of local stormwater management plans.” H.B. 987, 2012 Gen. Assemb.

Reg. Sess. (Md.2012); see also EN § 4-202.1(a)(1) (This “section applies to a county or municipality that is subject to a [NPDES Phase I MS4 permit].”). 35 MDE had investigated the costs of meeting the Bay TMDL and commissioned a study that revealed that “stormwater BMPs likely represent the largest costs to local governments in implementing the TMDL.” H.B. 987, 2012 Gen. Assemb.

Reg. Sess. (Md. 2012). The Permits also contain annual reporting requirements for: (1) the components of the stormwater management programs, and (2) data pertinent to the assessment of progress in implementing the Permit requirements, such as impervious surfaces and pollutant load reductions.

MDE will review the Counties’ reports to assess “progress toward meeting WLAs developed under EPA approved TMDLs” and the effectiveness of the programs in “reducing the discharge of pollutants to the MEP to protect water quality.” MDE will require BMP and program modifications if the Counties fail to comply with the Permit or show progress. The Permits also contain provisions setting forth sanctions for the violation of Permit conditions, including civil and criminal penalties. See, e.g., Montgomery County NPDES Permit Part VI.C (“Failure to comply with a permit provision ¹¹⁶ constitutes a violation of the CWA and is grounds for enforcement action; permit termination, revocation, or modification; or denial of a permit renewal application.”). Circuit Court and Court of Special Appeals Opinions EN § 1-601 provides for judicial review of MDE’s final determination to issue a permit. 36 The Water Groups 37 challenged the Permits in the various counties where MDE issued them.

The Circuit Court for Montgomery County remanded for MDE to revise the Permit in accordance with its opinion and order. In a reported opinion, the Court of Special Appeals affirmed. Md. Dep’t of the Env’t v. Anacostia Riverkeeper, 222 Md.App. 153, 157 , 112 A.3d 979, 981 (2015), cert. granted, 443 Md. 734 , 118 A.3d 861 (2015). MDE filed a petition for writ of certiorari, which we granted.

The Circuit Court for Baltimore County, the Circuit Court for Anne Arundel County, and the Circuit Court for Prince George’s County affirmed MDE’s decision to issue those Permits. The Water Groups filed notices of appeal to the Court of Special Appeals and, upon MDE’s motion, the Court of Special Appeals consolidated these three cases. MDE then filed a petition for writ of certiorari to this Court with questions nearly identical to those MDE submitted in its petition for writ of certiorari with respect to the Montgomery County Permit. Finally, the Circuit Court for Baltimore City also affirmed MDE’s decision to issue the Baltimore City Permit.

The ¹¹⁷ Water Groups filed a notice of appeal, and the Mayor & City Council of Baltimore (“Baltimore City”) filed a petition for writ of certiorari with a request that we consider this petition in conjunction with MDE’s petitions. We granted the City’s petition. As the Water Groups state in their brief, “the underlying Permits are substantively identical” and “are affected by the same legal defects.” We agree that the Permits are so substantively similar that we will analyze the agreed upon questions brought before the Court with respect to all the challenged Permits. We have slightly rephrased the questions: 1.

Did the MS4 permits issued by MDE for the counties’ municipal storm sewer system appropriately incorporate by reference publicly available materials and was the requirement for restoration of 20% of pre-2002 developed impervious surfaces specific, measurable, and enforceable? 2. Was MDE’s final decision to issue the permits with a 20% restoration requirement based upon the State’s Chesapeake Bay TMDL strategies, and a reporting requirement to establish strategies to address wasteload allocations, supported by substantial evidence? 3. Do the provisions of the MS4 permits that require that the public have an opportunity to review and comment on restoration plans intended to meet the wasteload allocations established for the permittees under applicable total maximum daily loads satisfy public participation requirements? 4. Do the provisions of the MS4 permits satisfy federal monitoring requirements?

We uphold MDE’s decision to issue the Permits on all grounds. Thus, we reverse the judgment of the Court of Special Appeals, which did not uphold the Montgomery County Permit, and we affirm the judgments of the circuit courts, which upheld the Permits in Anne Arundel County, Baltimore City, Baltimore County, and Prince George’s County. ¹¹⁸ STANDARD OF REVIEW Before 2009, challenges to the issuance or denial of a discharge permit were subject to a contested case hearing. Md.Code (1984, 2014 Repl.Vol.), § 10-222 of the State Government Article (“SG”), which is part of Maryland’s Administrative Procedure Act, delineates the grounds upon which a court can reverse an agency decision in a contested case. Specifically, SG § 10-222 provides that a court can reverse an agency decision in a contested case that “is unsupported by competent, material, and substantial evidence in light of the entire record as submitted” or that “is arbitrary or capricious.” In 2009, the General Assembly changed the procedures for challenging a discharge permit.

EN § 1-601 now permits direct judicial review of agency permitting decisions without a contested case hearing. Although this statute does not set forth a standard of review, the substantial evidence and arbitrary and capricious standards apply where an “organic statute” authorizes judicial review without a contested case hearing and does not set forth a standard of review. See Supervisor of Assessments of Carroll Cnty. v. Peter & John Radio Fellowship, Inc., 274 Md. 353, 355 , 335 A.2d 93, 94 (1975) (“Our cases have held that where no scope of review is thus provided, decisions of an administrative body will not be disturbed on appeal unless they are not supported by substantial evidence or are arbitrary, capricious or unreasonable.”) (citations omitted); Med. Waste Assocs., Inc. v. Md. Waste Coal., Inc., 327 Md. 596, 621 , 612 A.2d 241, 253 (1992) (“In an action for judicial review of an administrative decision granting a permit, however, the court determines not only whether the agency’s decision to issue the permit was in accordance with law, but also whether the particular administrative decision was arbitrary, capricious or unsupported by substantial evidence in light of the record as a whole.”).

Thus, even though all challenges going to the merits of the Permits in these consolidated cases originated in the Circuit Courts, 38 we ¹¹⁹ will review MDE’s decision to issue the Permits under the substantial evidence and arbitrary and capricious standards of review. Applying the substantial evidence standard of review to a case where no contested case hearing took place may seem anomalous because there is no formal record that was presented before an administrative law judge. EN § 1-606, however, expressly details the documents that can be included in a record. EN § 1-606(c)(1)-(9). 39 For example, EN § 1-606 stipulates that any draft permit, comments submitted to MDE ¹²⁰ during the public comment period, transcripts of public hearings on the permit application, and responses to submitted comments constitute part of the administrative record.

Thus, we are essentially reviewing the same record that we would have examined, excluding the administrative law judge’s decision, had the merits of this case been subject to a contested case proceeding. Accordingly, our review of the issuance of the Permits fits within the substantial evidence standard of review contemplated by SG § 10-222. In a review for substantial evidence, we ask “whether a reasoning mind reasonably could have reached the factual conclusion the agency reached.” Najafi v. Motor Vehicle Admin., 418 Md. 164, 173 , 12 A.3d 1255, 1261 (2011). We should accord deference “ ‘to the agency’s fact-finding and drawing of inferences’ ” when the record supports them.

Id. (citation omitted); see Mayor & Aldermen of City of Annapolis v. Annapolis Waterfront Co., 284 Md. 383, 399 , 396 A.2d 1080, 1089 (1979) (“The court may not substitute its judgment on the question whether the inference drawn is the right one or whether a different inference would be better supported. The test is reasonableness, not rightness.”) (citation and internal quotation marks omitted). Moreover, we shall review the agency’s decision “ ‘in the light most favorable to it.’ ” Najafi, 418 Md. at 173 , 12 A.3d at 1261 .

Finally, we must accord an agency great deference regarding factual questions involving scientific matters in its area of technical expertise. Bd. of Physician Quality Assurance v. Banks, 354 Md. 59, 69 , 729 A.2d 376, 381 (1999) (“[T]he expertise of the agency in its own field should be respected.”). We have characterized the arbitrary and capricious standard of review as one that is “extremely deferential.” Harvey v. Marshall, 389 Md. 243, 299 , 884 A.2d 1171, 1205 (2005). In reviewing the issuance of an NPDES permit, the U.S. Court of Appeals for the Second Circuit quoted language derived from Motor Vehicle Mfrs.

Ass’n v. State Farm Mut. Auto. Ins. Co., 463 U.S. 29, 43 , 103 S.Ct. 2856 , 77 L.Ed.2d 443 (1983) — the ¹²¹ United States Supreme Court’s leading case on the arbitrary and capricious standard: To determine whether the agency’s actions were “arbitrary and capricious,” we consider whether the agency ‘relied on factors which Congress has not intended it to consider, entirely failed to consider an important aspect of the problem, offered an explanation for its decision that runs counter to the evidence before the agency, or is so implausible that it could not be ascribed to a difference in view or the product of agency expertise.’ Natural Res.

Def. Council, 808 F.3d at 569 (citations and quotation marks omitted). The court also elaborated that “[w]e must be ‘satisfied from the record that the agency ... examine[d] the relevant data and articulate[d] a satisfactory explanation for its action’ ” and that it “afford[ed] the agency’s decision greater deference regarding factual questions involving scientific matters in its area of technical expertise.” Id. (quoting State Farm, 463 U.S. at 43 , 103 S.Ct. 2856 ) (citations and internal quotation marks omitted).

The Second Circuit’s articulation of the arbitrary and capricious standard is in accord with Maryland’s treatment of this standard as one that is highly deferential. See Harvey, 389 Md. at 299 , 884 A.2d at 1205 . We are therefore mindful of the Second Circuit’s explanation of the principles underlying the arbitrary and capricious standard when applying that standard to this case. 40 ¹²² In addition, we will review an agency decision for an error of law. When our review concerns a legal question, we apply less deference to the agency’s conclusions.

HNS Dev., LLC v. People’s Counsel for Balt Cnty., 425 Md. 436, 449 , 42 A.3d 12, 20 (2012). We refuse to uphold an agency decision “premised solely upon an erroneous conclusion of law.” Id. (citation omitted) (internal quotation marks omitted). Otherwise, we ordinarily should give “considerable weight” to an agency’s “interpretation and application of the statute which [it] administers.” W.R. Grace & Co. v. Swedo, 439 Md. 441, 453 , 96 A.3d 210, 217 (2014); John A. v. Bd. of Educ. for Howard Cnty., 400 Md. 363, 381-82 , 929 A.2d 136, 147 (2007) (“In reviewing an agency’s legal conclusions, it is a fundamental principle of administrative law that a reviewing court should not substitute its judgment for the expertise of those persons who constitute the administrative agency.”).

DISCUSSION Part I: The 20% Restoration Requirement The Permits require, by the end of the five-year term, that the Counties restore 20% of the impervious surface areas in their watersheds that have not been restored to the MEP. This requirement “uses percent impervious cover in a watershed as a surrogate TMDL target.” ENSR, Pilot TMDL Applications Using the Impervious Cover Method § 1.0, at 1-1 (2005). Like so much of this case, we must unpack the science before we analyze the parties’ arguments. As we develop on land, science has shown us that we profoundly impact our waters.

Consider, for example, when “[t]rees, meadow grasses, and agricultural crops that had intercepted and absorbed rainfall are removed.... ” CWP & MDE, Manual, § 1.1, at 1.3. Problematically, “[c]leared and graded sites erode, are often severely compacted, and can no longer prevent rainfall from being rapidly converted into stormwater runoff.” Id. These kinds of sites are known as impervious surfaces, surfaces “that do[ ] not allow stormwater to infiltrate into the ground,” such as “rooftops, driveways, ¹²³ sidewalks, or pavement.” EN § 4-201.1(d). “Impervious surfaces accumulate pollutants deposited from the atmosphere,” pollutants which are “rapidly delivered to downstream waters” during storms. CWP & MDE, Manual, § 1.1.1, at 1.5.

The purpose of the 20% restoration requirement, then, is to use stormwater management practices to restore the natural, beneficial processes in our environment that we have changed by developing impervious surfaces. In other words, the 20% restoration requirement is a surrogate because the requirement does not control pollution reduction directly. See ENSR, Pilot TMDL Applications Using the Impervious Cover Method § 1.0, at 1-1. Rather, it is through restoring impervious surfaces with management practices that the Counties will reduce pollution.

See, e.g., CWP & MDE, Manual, § 1.2, at 1.13 (“[Management practices] shall be designed to remove 80% of the average annual post development total suspended solids load (TSS) and 40% of the average annual post development total phosphorus load (TP).”). A. Maximum Extent Practicable The Water Groups argue that the 20% restoration requirement is too opaque to comply with 33 U.S.C. § 1342 (p)(3)(B)(iii), the MEP standard. They so argue because, they contend, MDE “failed to provide a specific performance standard for restoration activities” or a “numeric limitation ... for what pollution reductions must be accomplished by the permittees’ twenty-percent restoration efforts.” They also argue that MDE failed to explain what impervious surface is “not restored to the MEP.” We disagree because (1) the applicable law does not impose a specific performance standard on MS4s and (2) MDE did actually select a performance standard for the Counties to adhere to. 33 U.S.C. § 1342 (p)(3)(B)(iii) states: Permits for discharges from municipal storm sewers shall require controls to reduce the discharge of pollutants to the maximum extent practicable, including management practices, control techniques and system, design and engineering methods, and such other provisions as the Administrator or ¹²⁴ the State determines appropriate for the control of such pollutants. From the text, we discern that Congress established a broad requirement for MS4s.

The list of required controls is not exclusive. See 55 Fed.Reg. at 48,038 (“[MS4] controls may be different in different permits.”). And the purpose of the controls — reducing the discharge of pollutants — exists alongside the flexible, undefined standard “to the maximum extent practicable.” See City of Abilene v. EPA, 325 F.3d 657, 659-60 (5th Cir.2003); Natural Res. Def.

Council, Inc. v. N.Y. State Dep’t of Envtl. Conservation, 25 N.Y.3d 373, 406 , 13 N.Y.S.3d 272 , 34 N.E.3d 782 (N.Y.2015) (Rivera, J., dissenting in part) (“The CWA does not define the maximum extent practicable standard. However, it appears to provide broad authority to agencies to control stormwater pollution.”); 55 Fed.Reg. at 48,038 (“In enacting section 405 of the WQA [Water Quality Act], Congress recognized that permit requirements for [MS4s] should be developed in a flexible manner to allow site-specific permit conditions to reflect the wide range of impacts that can be associated with these discharges.”). 33 U.S.C. § 1342 (p)(3)(B)(iii) imposes no minimum standard or requirement on MDE other than to establish controls for MS4s to reduce the discharge of pollutants. See Natural Res.

Def. Council, Inc. v. EPA, 966 F.2d 1292, 1308 (9th Cir.1992) (“Congress did not mandate [in § 1342(p)(3)(B)(iii) ] a minimum standards approach or specify that EPA develop minimal performance requirements.....Congress could have written a statute requiring stricter standards, and it did not.”). Thus, we reject the Water Groups’ argument that MDE committed legal error by “failing] to provide a specific performance standard for restoration activities.” 41 ¹²⁵ Moreover, MDE tethered restoration to the practices in the Manual, which MDE incorporates into the Permits by reference. See Part III.E (“These management programs are designed to control stormwater discharges to the maximum extent practicable____”); Part III.E.1 (“At a minimum, the County shall ... [i]mplement the stormwater management ... practices found in the [Manual]____”).

The Manual explains that the list of acceptable stormwater management practices is tied to the WQ,. “The Water Quality Volume (denoted as the WQy) is the storage needed to capture and treat the runoff from 90% of the average annual rainfall.” CWP & MDE, Manual, § 2.1, at 2.2. 42 The Manual further explains that the “WQV is directly related to the amount of impervious cover created at a site.” Id. In other words, MDE chose a standard that relates to the very problem the 20% restoration requirement serves to abate: the increase in stormwater runoff and the discharge of pollutants because of the increase in impervious surfaces. See CWP & MDE, Manual, § 1.1, at 1.4 (“As can be seen, the volume of stormwater runoff increases sharply with impervious cover”). Thus, the record reflects that MDE has established a performance standard, WQV, that defines as acceptable those practices the Counties may choose from to fulfill the 20% restoration requirement.

See id. § 2.7 (Acceptable Urban BMP Options). ¹²⁶ Moreover, our discussion of restoration is instructive as to why, despite the Water Groups’ contention, the “impervious surface area that is not restored to the MEP” is sufficiently clear and measurable. (Emphasis added.) The area that is not restored to the MEP is the area without the restoration controls described in the Manual. Moreover, the Manual explains that impervious area refers to an area “that does not have vegetative or permeable cover.” CWP & MDE, Manual, § 2.1, at 2.4. Put together, the “impervious surface area that is not restored to the MEP” refers to a defined type of area (impervious surface) lacking a type of stormwater management control (the BMPs in the Manual).

By way of example, previous MS4 reports delineate these criteria so that MDE can evaluate whether Montgomery County (in this example) installed the required controls. See Montgomery County Department of Environmental Protection, Annual Report for 2006 NPDES MS4 Permit F2, at III-64; see also 33 U.S.C. § 1342 (p)(3)(B)(iii) (MS4 permits “shall require controls” such as management practices.). Because 33 U.S.C. § 1342 (p)(3)(B)(iii) does not require a specific performance standard, and because the concepts of restoration and impervious surface “not restored to the MEP” are sufficiently clear as to the controls that the Counties must install, the 20% restoration requirement in the Permits complies with the MEP standard. See 33 U.S.C. § 1342 (p)(3)(B)(iii).

B. Substantial Evidence and Arbitrary and Capricious The Water Groups also argue that MDE has not explained why it selected 20% as the restoration goal or how this Permit provision will promote necessary pollution reduction. The Water Groups contend that MDE ineffectively justifies its choice based on the Bay TMDL because the Permits do not assure that the Counties will achieve the Bay TMDL’s objectives or reductions. Even accepting a connection between the 20% restoration requirement and the Bay TMDL, the Water Groups argue that MDE still failed to take ¹²⁷ into account numerous other TMDLs related to the Counties’ waters. We disagree with the Water Groups’ position because (1) the applicable law affords permitting authorities flexibility in establishing controls for MS4s and (2) MDE has justified its decision based on a well-developed and vetted strategy.

Natural Res. Def. Council, 808 F.3d at 569 (citation and internal quotation marks omitted). (“We must be ‘satisfied from the record that the agency ... examine[d] the relevant data and articulate[d] a satisfactory explanation for its action.’ ”).

Congress established a flexible framework in 33 U.S.C. § 1342 (p)(3)(B)(iii). As the text states, MS4s shall require controls “and such other provisions as the Administrator or the State determines appropriate for the control of such pollutants.” 33 U.S.C. § 1342 (p)(3)(B)(iii) (emphasis added). MDE has determined that the 20% restoration requirement is a key strategy in restoring the Chesapeake Bay. See Part VI.A (Chesapeake Bay Restoration by 2025.).

The Permits in particular state that the Counties will carry out the 20% restoration requirement in order to comply with the Bay TMDL. Id. A review of Maryland’s Watershed Implementation Plan (“WIP”) is instructive as to why the 20% restoration requirement will help to restore the Chesapeake Bay. MDE, Basis for Final Determination.

As we have discussed, WIPs are “road-maps” setting forth a plan for how and when a jurisdiction will reach the pollution reduction goals in the Bay TMDL. EPA, Bay TMDL ES at ES-8. The EPA developed the Bay TMDL “in reliance on” the WIPs that each affected jurisdiction submitted to the EPA. Am.

Farm Bureau Fed’n, 792 F.3d at 291 (emphasis added). Moreover, the EPA approved the Bay TMDL “only after” determining that each jurisdiction provided “reasonable assurance” that it would meet the pollutant reductions in its WIP. Id. (emphasis added); see EPA, Bay TMDL ES at ES-8 (The WIPs are the “cornerstone” that ¹²⁸ ensures accountability to achieve pollution reductions). 43 Importantly, the strategies in Maryland’s WIP for urban stormwater include, among other things, the reduction of impervious surface area by 20% over a five-year period, just as the Permits do.

Compare Maryland Phase I Watershed Implementation Plan: Executive Summary ES-15 (Dec. 3, 2010) [hereinafter Phase I WIP ES], with Part III.G.2. In fact, restoration of “twenty percent of the counties’ impervious surface area that is not already restored to the maximum extent practicable (MEP)” is listed in the “key elements” that provide reasonable assurance of the implementation of Maryland’s WIP. Maryland Phase I Watershed Implementation Plan: § 5.2.2, at 5-30 (Dec. 3, 2010). MDE incorporated another important element from the WIP to bolster compliance with the 20% restoration requirement.

The EPA explained that the WIPs were to include “mechanisms to track and report implementation activities” and “provide alternative approaches.” EPA, Chesapeake Bay TMDL ES at ES-8-ES-9. Maryland’s WIP thus includes what we refer to as the adaptive management approach, whereby additional or alternative practices are implemented if existing programs are not meeting target reductions. Phase I WIP at 5-30. 44 MDE incorporated this approach from the WIP into the Permits as part of the agency’s review of the Counties’ annual reports: “BMP and program modifications shall be made” if the Counties fail to comply with the Permits or fail to show progress in meeting WLAs of EPA-approved TMDLs. 45 We are satisfied from the record that MDE “ex ¹²⁹ amine[d] the relevant data and artieulate[d] a satisfactory explanation for its action.” Natural Res. Def.

Council, 808 F.3d at 569 (citations and internal quotation marks omitted). We conclude, then, that MDE’s decision to include the 20% restoration requirement in the Permits was supported by substantial evidence and is not arbitrary and capricious. 46 Although we recognize that MDE issued the Montgomery County Permit before Maryland’s WIP and the Bay TMDL were prepared, the Bay TMDL was not an isolated event. See Am. Farm Bureau, 984 F.Supp.2d at 298 (The Bay TMDL “is not a new or recent idea,” and thus, “it would be improper to view the Final TMDL in a vacuum as a single, isolated effort to restore water quality to the Chesapeake Bay.”).

As we have discussed, an important prior effort to restore the Chesapeake Bay was the Chesapeake 2000 Agreement. To achieve the restoration goals of that agreement, the Governor of Maryland’s Chesapeake Bay Cabinet prepared a draft of the State’s plan. Department of Legislative Services, Office of Policy Analysis, Chesapeake Bay Restoration and the Tributary Strategy: An Analysis of Maryland’s Efforts to Meet the Nutrient and Sediment Reduction Goals of the Chesapeake 2000 Agreement iii (2007). The strategy for stormwater included restoration of impervious surface in Montgomery County, the first County to successfully implement this program.

Id. at 13. The State conceived of this strategy as an effective ¹³⁰ state-wide method of improving the Chesapeake Bay. See Maryland’s Chesapeake Bay Tributary Strategy Statewide Implementation Plan 16 (2008) (stormwater strategy). Thus, MDE’s decision to include the 20% restoration requirement in the Montgomery County Permit was also supported by substantial evidence and was not arbitrary and capricious.

We further disagree with the Water Groups’ position that the 20% restoration requirement is insufficient because it does not relate to other TMDLs. Indeed, the Permits incorporate a requirement to submit plans regarding WLAs for all EPA-approved TMDLs. That requirement ensures the Permits address all applicable TMDLs. 47 C. The Methodology in the Guidance Finally, with respect to the 20% restoration requirement, the Water Groups object to MDE’s method of calculating impervious surface area not restored to the MEP. They assert that the Guidance, which MDE incorporated into these Permits by reference, is flawed because MDE arbitrarily selected 2002 as the baseline for measuring impervious surface area.

For example, the Water Groups allege that MDE has ignored the current definition for MEP in the agency’s own regulations. We disagree with the Water Groups because MDE reasonably justifies its decision based on the accurate determination that 2002 marked a significant milestone in the State’s treatment of water quality. See Najafi, 418 Md. at 173 , 12 A.3d at 1261 (discussing the substantial evidence test); Natural Res. ¹³¹ Def. Council, 808 F.3d at 569 (discussing the arbitrary and capricious test).

MDE explained its selection of 2002 in the Guidance: “Maryland regulations and local ordinances began requiring BMPs [around this time] to address a specific suite of volumes [recharge (Rey), water quality (WQy), and channel protection (Cpv) ] and it can therefore be justified that water quality treatment has been provided to the MEP.” MDE, Guidance at 4. In proposing new regulations for stormwater management in 2000, MDE explained: “The resulting program will provide water quality treatment of up to 90 percent of the average annual rainfall throughout the State, establish ground water recharge standards, and outline a channel erosion control strategy.” 27 Md. Reg. at 1168. Additionally, MDE explained that prior regulations had provided “sparse guidance” on “water quality enhancement.” Id. To bolster MDE’s focus on water quality, the agency incorporated the Manual by reference into the 2000 regulations.

See id. at 1170; see also CWP & MDE, Manual § 2.1, at 2.2 (“The WQy is directly related to the amount of impervious cover created at a site.”). The Water Groups reject MDE’s decision by arguing that Maryland’s regulations currently define “maximum extent practicable” based on a different requirement the State imposed after 2002. That is, COMAR 26.17.02.02B(22) defines MEP as the design of “stormwater management systems so that all reasonable opportunities for using ESD planning techniques and treatment practices are exhausted and, only where absolutely necessary, a structural BMP is implemented.” (Emphasis added.) MDE has explained, however, that pursuant to the Storm-water Management Act of 2007, which precipitated the regulatory change the Water Groups refer to, 48 the performance requirement ESD pertains to “future development,” not to restoration. MDE, Basis for Final Determination.

As set forth in EN § 4-203(b)(1), (“[t]he rules and regulations shall ¹³² [i]ndicate that the primary goal of the State and local programs will be to maintain after development, as nearly as possible, the predevelopment runoff characteristics.”) (emphasis added). See also COMAR 26.17.02.01B (“These regulations for stormwater management apply to the development or redevelopment of land____”). MDE has reached a correct legal conclusion, and one regarding the law in which the agency has expertise. See John A., 400 Md. at 381-82 , 929 A.2d at 147 (“In reviewing an agency’s legal conclusions, it is a fundamental principle of administrative law that a reviewing court should not substitute its judgment for the expertise of those persons who constitute the administrative agency.”).

We find no error here. Although the agency selected ESD as the MEP via regulation in 2009, 49 MDE chose 2002 because that year marked the beginning of the stormwater management era when BMPs were subjected to performance standards to better treat water quality. 27 Md. Reg. at 1168. WQv is one such standard. CWP & MDE, Manual § 1.2, at 1.13.

The Manual shows a direct relationship between the amount of impervious surface and WQv. Id. § 2.1, at 2.3 (Figure 2.2). Thus, beginning in 2002, Maryland connected stormwater management practices to the restoration of impervious surface area, which is the Permit requirement at issue. MDE’s conclusion that the Counties should calculate impervious surface area using 2002 as a baseline as part of the 20% restoration requirement was supported by substantial evidence and was not arbitrary and capricious.

See Banks, 354 Md. at 69 , 729 A.2d at 381 (“[T]he expertise of the agency in its own field should be respected.”). 50 ¹³³ Because we uphold the Guidance as a component the Counties may legally use to achieve the 20% restoration requirement, we must address whether Montgomery County can use this methodology, too. We recognize that MDE issued the Montgomery County Permit before MDE prepared the Guidance. Compare Montgomery County Permit: Part I: C (Effective Date: February 16, 2010), with MDE, Basis for Final Determination at 10 (MDE held meetings with the Counties about the Guidance from August 2010 to February 2011 before completing it in June 2011.). As we have discussed, MDE limited the practices the County could implement by connecting restoration to the Manual.

But, importantly, MDE did not limit the manner in which Montgomery County could show its compliance with the 20% restoration requirement. See, e.g., Montgomery County Permit: Part G.2 (Watershed Restoration), Part IV.A (Annual Reporting). Because the Permit does not restrict how the County must show its progress in achieving the 20% restoration requirement and because of the important links we have already analyzed between the Guidance and the Manual (i.e., the WQy standard), we conclude that Montgomery County can rely on the Guidance in showing its compliance with this requirement. Part II: TMDL Plans As we discussed at the outset of this opinion, TMDLs are informational tools, of which WLAs — -wasteload allocations — are critical.

As the District Court for the Middle District of Pennsylvania aptly noted, “WLAs are not permit limits per se; rather they still require translation into permit limits — ” Am. Farm Bureau Fed’n, 984 F.Supp.2d at 328 (citation and internal quotation marks omitted) (emphasis in original). The EPA requires such translation pursuant to 40 C.F.R. § 122.44 (d)(1)(vii)(B), which the Water Groups argue MDE failed to comply with. 51 ¹³⁴ Pursuant to 40 C.F.R. § 122.44 (d)(1)(vii)(B), MDE included a provision that requires the Counties to submit plans for all EPA-approved TMDLs one year after the issuance of the Permits (or alternatively, one year after the EPA approves the TMDL). 52 In creating these restoration plans, the Counties must select actions (including cost estimates) and set forth a schedule (including deadlines and pollution reduction benchmarks) to meet WLAs. Because MDE foresaw the process of meeting WLAs as “iterative,” the Counties must also describe what actions they will take when they fail to make progress in meeting WLAs. 53 Effluent Limitations Must Be Consistent with WLAs 40 C.F.R. § 122.44 (d)(1)(vii)(B) requires MDE to establish effluent limitations that take into account WLAs: When developing water quality-based effluent limits under this paragraph the permitting authority shall ensure thatt ] ... [elffluent limits developed to protect a narrative water quality criterion, a numeric water quality criterion, or both, are consistent with the assumptions and requirements of any available wasteload allocation for ¹³⁵ the discharge prepared by the State and approved by EPA pursuant to 40 CFR 130.7.

(emphasis added). The Water Groups construe the contents of the TMDL implementation plans, that is, the actions and deadlines discussed supra, as effluent limitations. As they put it, nothing else in the Permits is “remotely capable of ... impos[ing] effluent limitations that are consistent with the assumptions and requirements of TMDLs.” Thus, the Permits can only be in compliance with 40 C.F.R. § 122.44 (d)(1)(vii)(B) when MDE has reviewed the contents of the TMDL implementation plans. The problem for MDE, then, is one of timing.

Because MDE will review the contents one year after issuing the Permits, the Water Groups conclude that MDE cannot know whether the effluent limitations in the Permits are consistent with the WLAs. We disagree with the Water Groups because (1) 40 C.F.R. § 122.44 (d)(1)(vii)(B) is, like the MEP standard, flexible as to how a permitting authority complies with this regulation and (2) MDE has complied with the regulation by incorporating the WLAs (the most critical element of the regulation) into the Permits and by using an “iterative” process of agency review and program change to ensure progress in meeting the WLAs. Before we analyze MDE’s “iterative” process, we must take issue with the Water Groups’ position that the contents of the restoration plans are effluent limits. We understand the effluent limits to be best management practices (“BMPs”). 54 But, importantly, the Permits do not incorporate these effluent limits in the restoration plans but in another section of ¹³⁶ the Permits — Part III.E. That is, the Counties must implement a stormwater management program (“SWMP”), including the “practices found in the 2000 Maryland, Stormwater Design Manual.” 55 Importantly, the Permits instruct that “these [management] programs shall be integrated with other permit requirements to promote a comprehensive adaptive approach toward solving water quality problems.” In this light, we understand the TMDL implementation plans as roadmaps because they describe a plan for achieving a goal (using BMPs to meet WLAs).

See, e.g., Montgomery County Permit Part III.J (In the TMDL implementation plans, the County must “describe those actions necessary to meet the storm drain system’s share of WLAs in EPA approved TMDLs.”). The requirement to implement BMPs, however, exists independent of the requirement to submit the TMDL implementation plans. Having clarified this matter, we turn to the regulation to better understand MDE’s obligation with respect to effluent limits and WLAs. We note that the plain text imposes a duty on the permitting authority to ensure that effluent limits are consistent with WLA assumptions and requirements.

See In re City of Moscow, Idaho, 2001 WL 988721 , at ⁹ (“[T]he governing regulations require consistency....”) (emphasis in original). But the text does not instruct the permitting authority as to how it must ensure this consistency. See id. at ⁸ (noting the lack of detail in the regulation). Instead, the EPA set a minimal, flexible requirement in which the permitting authority is to design a scheme where effluent limits are compatible or in agreement with WLAs.

See National Pollution Discharge Elimination System, 54 Fed.Reg. 23,868, 23,870, 23,879 (June 2, 1989) (codified at 40 C.F.R. § 122.44 ) (“Subparagraph (vii) does not prescribe detailed procedures ¹³⁷ for developing water quality-based effluent limits. Rather, the regulation prescribes minimum requirements for developing water quality-based effluent limits, and at the same time, gives the permitting authority the flexibility to determine the appropriate procedures for developing water quality-based effluent limits.”); The American Heritage Dictionary of the English Language 392 (4th ed. 2006) (Consistent means in agreement or compatible.) 56 ; cf. Am. Farm Bureau Fed’n, 984 F.Supp.2d at 328 (“Accordingly, in some circumstances, a state may write a NPDES permit limit that is different from the WLA, provided that it is consistent with the operative assumptions underlying the WLA”). In re City of Moscow, Idaho is illustrative of the flexibility the regulation affords MDE.

There, the Environmental Appeals Board rejected the City of Moscow’s petition for review of an EPA-issued NPDES permit. 2001 WL 988721 , at ¹. In pertinent part, the Board concluded that the EPA (Region X) did not err in creating permit limits although the EPA did not incorporate the design flow rate of an applicable TMDL. Id. at ⁸-9 . After reviewing 40 C.F.R. § 122.44 (d)(1)(vii)(B) and the regulatory history, see 54 Fed.Reg. 23,868, 23,879, the Board explained that there was no law or rule prescribing how the EPA was to select a flow rate to create effluent limits, In re City of Moscow, Idaho, 2001 WL 988721 , at ⁸.

The Board concluded that the agency acted “well within the discretion accorded [it] under the applicable regulatory scheme.” Id. at ⁹ . So too here. No law or regulation specifies how or burdens MDE in undertaking the process of complying with 40 C.F.R. § 122.44 (d)(1)(vii)(B). In fact, as we have discussed, the overarching federal law for MS4s — 33 U.S.C. § 1342 (p)(3)(B)(iii)— is broad and flexible, too.

See Natural Res. Def. Council, 966 F.2d at 1308 . ¹³⁸ Under the Permits, the Counties must include in the TMDL implementation plans the actions needed to meet “WLAs in EPA approved TMDLs.” Part III.J.1. Thus, the Permit ties the activities to WLAs, not to some other benchmark that could be inconsistent with the WLA.

As the Permits make clear, “[a]ll EPA approved TMDL’s that establish WLA’s applicable to the County’s storm drain system are incorporated by reference into this permit.” Id. The Counties must also submit “documentation of progress toward meeting applicable WLAs developed under EPA approved TMDLs.” Reporting on WLA progress is reinforced through watershed assessment and restoration requirements. See Part III.F (“[Watershed assessments shall [sjpecify how restoration efforts will increase progress toward meeting any applicable WLAs included in EPA approved TMDLs.”); Part III.G (The Counties shall annually report on progress toward meeting WLAs.). Moreover, “MDE will review program implementation, annual reports, and periodic data submittal on an annual basis.” The Permit requires “BMP and program modifications” if this report fails to “show progress toward meeting WLAs.” The modifications are a product of the “iterative approach” the Counties must follow if they do not make progress in achieving WLAs.

That is, the Counties must propose “additional or alternate stormwater controls” to meet WLAs, which MDE will review and approve, if adequate. Environmental Defense Center, Inc. v. United States Environmental Protection Agency (“EDC”) is instructive on the importance of permittee reporting and agency review to ensure compliance with the law. 344 F.3d 832 , 858 (9th Cir.2003). In EDC, the Ninth Circuit vacated a portion of an EPA rule that permitted small MS4s to discharge under a general permitting scheme, and remanded for the EPA to revise the rule to comply with the CWA. Id. at 858. 57 The ¹³⁹ Ninth Circuit stated that the rule did not require the permitting authority to review the MS4 operator’s stormwater management program “to ensure” that the operator’s program would “in fact” comply with the relevant law.

Id. at 855 (emphasis in original). Thus, the Ninth Circuit concluded that “nothing prevents the operator of a small MS4 from misunderstanding or misrepresenting its own stormwater situation.” Id. The scheme the Ninth Circuit rejected in EDC is materially different from the scheme in the Permits here. MDE requires reporting, assessment, and adaptation to ensure that the Counties’ BMPs will make progress to achieve WLAs.

We find the effluent limits under this scheme to be consistent with approved WLAs. See id. at 856 (“However, stormwater management programs that are designed by regulated parties must, in every instance, be subject to meaningful review by an appropriate regulating entity....”). Thus, the Permits comply with 40 C.F.R. § 122.44 (d)(1)(vii)(B). Part III: Monitoring Here, the Water Groups raise two arguments: the Permits do not include federally mandated monitoring to (1) produce representative data in the MS4 jurisdictions and (2) assure compliance with Permit requirements.

They contend that MDE elected not to comply with federal regulations that impose the above requirements and that the record does not otherwise show that MDE complied with these laws. MDE responds by defending its monitoring scheme as sufficient to comply with the applicable federal regulations. Moreover, MDE argues that some of the federal provisions are inapposite because they only pertained to the initial application for MS4 permits in the 1990s. ¹⁴⁰ A. The Permit’s Monitoring Requirements We begin by noting what the Permits require. Under Part III.F, each County must conduct chemical, biological, and physical monitoring at one outfall and an associated in-stream station. 58 The Permits set forth the number of required monitoring events, sampling methods, pollutants, and locations. 59 The Counties must also conduct monitoring of stream channel protection. 60 These two levels of monitoring occur at different locations.

Compare note 58, with note 60. ¹⁴¹ B. Applicable Law We now consider what federal monitoring regulations apply to the Permits. Under 40 C.F.R. § 122.48 (b), “[ajll permits shall specify [Required monitoring including type, intervals, and frequency sufficient to yield data which are representative of the monitored activity including, when appropriate, continuous monitoring.” (Emphasis added.) Moreover, monitoring requirements must “assure compliance with permit limitations.” 40 C.F.R. § 122.44 (i)(1); see also In re Gov’t of the D.C. Mun. Separate Storm Sewer Sys., 10 E.A.D. 323 , 2002 WL 257698 , at ²⁰ (Feb. 20, 2002) (“First, both section 122.48(b) and section 122.44(i) would appear to require that certain monitoring conditions be included in all permits.”). The Clean Water Act makes plain the EPA’s broad authority to set these requirements. 33 U.S.C. § 1318 (a); Natural Res.

Def. Council, Inc. v. EPA, 863 F.2d 1420 , 1434 (9th Cir.1988) (The EPA “has wide discretion and authority to determine monitoring requirements in NPDES permits____”). These requirements apply to state programs. 40 C.F.R. § 123.25 (15), (19). The Water Groups point us to an additional regulation as part of their argument that the Permits are not capable of producing representative data.

Specifically, pursuant to 40 C.F.R. § 122.26 (d)(2)(iii)(D), applicants for large MS4s must submit a “proposed monitoring program for representative data collection for the term of the permit that describes the location of outfalls or field screening points to be sampled (or the location of instream stations),” and, among other things, “why the location is representative.” But MDE disputes whether this regulation is even applicable to our analysis of the agency’s decision to issue the Permits when the regulation, in its view, pertains to applicants (i.e., permittees). MDE argues that we should read 40 C.F.R. § 122.26 (d)(2)(iii)(D) as relevant only to a permit application — not a permit — and, moreover, only to the first cycle of permit applications for MS4s (i.e., the early 1990s). In MDE’s view, the agency had “the flexibility to establish requirements [it] deem[ed] appropriate” in subsequent MS4 permits. ¹⁴² Recognizing that 40 C.F.R. § 122.26 (d) concerns “[application requirements,” we nevertheless agree with the Water Groups that the regulation is relevant to our review of MDE’s decision to issue the Permits. Although state and federal permitting laws are unhelpful, an EPA Policy Memorandum interpreting 40 C.F.R. § 122.26 (d) (“the Policy Memo”) 61 is instructive.

In the Policy Memo, the EPA instructs that “reapplications should focus on maintenance and improvement of [MS4 storm water management] programs” that were required in the initial applications. EPA, Interpretive Policy Memorandum on Reapplication Requirements for Municipal Separate Sewer Systems, 61 Fed.Reg. 41,698, 41,698 (Aug. 9, 1996) [hereinafter Policy Memo]. 62 According to the Policy Memo, MDE can make changes to the MS4 monitoring program during the reapplication period, , but such changes must be “appropriate and useful.” Id. at 41,699 . In other words, when a permittee reapplies for a discharge permit, and if the ¹⁴³ permitting authority reissues the permit, it is the permitting authority’s responsibility to ensure that the reissued permit contains programs that are adequate in light of the initial application requirements in 40 C.F.R. § 122.26 (d). We rely, then, on 40 C.F.R. § 122.26 (d) as a baseline to review the monitoring program in the Permits to ensure that MDE only made program changes that were “appropriate and useful.” Id. at 41,699 .

C. The EPA Policy Memo The EPA published the Policy Memo to “respond[ ] to requests from municipalities and NPDES permit writers for clarification about regulations which do not appear to address reapplication requirements, i.e., permit reissuance.” 61 Fed. Reg. at 41,698. In the Policy Memo, the EPA first explains that it does not require an applicant to repeat in full the process in 40 C.F.R. § 122.26 (d)(1)-(2). Id. The EPA explained that it would be redundant to “request the same information again [including characterization of data at § 122.26(d)(2)(iii) ], where it has already been provided and has not changed.” Id.

But an applicant “should identify any proposed changes or improvements” to “monitoring activities.” Id. If MS4 applicants and permit writers wish to change programs in future permits, the EPA explains, they may not only “de-emphasize” but also propose to eliminate a program component. Id. at 41,699; see id. at 41,698 (“MS4 permit applicants and NPDES permit writers have considerable discretion to customize appropriate and streamlined reapplication requirements on a case-by-case basis.”). De-emphasis may be justified based on the discharger’s experience during the first permit term.

Id. By way of example, the EPA stated that “new information on the relative magnitude of a problem” and “new data on water quality impacts of storm water discharges” could justify program changes. Id. Moreover, elimination could be justified “when a different water quality program would serve the same goals.” Id.

We construe the phrase “same goals” as referring to the CWA’s objective in Chapter ¹⁴⁴ 26 of Title 33 of “restoring] and maintaining] the chemical, physical, and biological integrity of the Nation’s waters.” 33 U.S.C. § 1251 (a); see 33 U.S.C. § 1318 (a)(1)(A)(i) (“Whenever required to carry out the objective of this chapter [26: Water Pollution Prevention and Control],____the Administrator shall require the owner or operator of any point source to establish and maintain such records ... [and use such monitoring methods] ... as he may reasonably require.”) (emphasis added). If MS4 applicants and permit writers wish to change monitoring programs, the EPA encouraged them “to determine if storm water monitoring efforts are appropriate and useful.” 61 Fed.Reg. 41,699. The EPA then suggested alternative techniques: “municipalities may wish to consider using monitoring techniques other than end-of-the pipe chemical-specific monitoring, including habitat assessments, bioassessments, and/or other biological methods.” Id. (emphasis added).

D. Representative Monitoring MDE argues, somewhat correctly, that the Policy Memo grants it “the flexibility to establish requirements [it] deem[s] appropriate.” The Water Groups only repeat the language of the Policy Memo, arguing that MDE can change programs to make them “more appropriate and useful.” They argue that MDE’s decision was not supported by substantial evidence. Although the Policy Memo does grant NPDES permit writers, like MDE, the flexibility to modify program components, including monitoring, 61 Fed.Reg. at 41,699, the EPA does not allow permit writers to reissue permits and abdicate their responsibility in “restor[ing] and maintaining] the chemical, physical, and biological integrity of the Nation’s waters.” 33 U.S.C. § 1251 (a). 40 C.F.R. 122.26(d) required, among other things, a “proposed monitoring program for representative data collection for the term of the permit that describes the location of outfalls or field screening points to be sampled (or the location of instream stations),” and, among other things, “why the location is representative.” And as we also discussed earlier, the Permits must include monitoring “sufficient ¹⁴⁵ to yield data which are representative of the monitored activity.” 40 C.F.R. § 122.48 (b). We conclude that MDE’s monitoring program in the Permits will produce representative data because MDE has (1) ensured that the Counties monitor stormwater discharges at monitoring locations that represent an adequate range of land uses statewide, and (2) increased the frequency of monitoring to yield more representative information at the County level. MDE revised the MS4 monitoring program in the late 1990s to implement a monitoring approach called the “three legged stool,” so named because it incorporated not only “[c]hemical testing” but also “biological criteria” and “physical assessments.” MDE, Maryland’s NPDES Municipal Stormwater Monitoring at § 4.2 (1997); see also MDE, Basis for Final Determination to Issue Anne Arundel County’s NPDES MS4 Permit.

The agency articulated this approach as being “more aligned with the CWA’s goal to ‘restore and maintain the chemical, physical, and biological integrity of the nation’s waters.’ ” MDE, Maryland’s NPDES Municipal Stormwater Monitoring at § 4.2 (emphasis added); see also MDE, Basis for Final Determination to Issue Anne Arundel County’s NPDES MS4 Permit. MDE also increased the frequency of monitoring so that the Counties needed to monitor twelve storm events instead of three. MDE, Maryland’s NPDES Municipal Stormwater Monitoring at § 4.0. Finally, in light of the increase in monitoring type, MDE decreased the number of monitoring sites to one outfall and one associated in-stream station in a watershed.

Id. at § 4.3. The agency explained that it adopted the “three legged stool” approach after an analysis of the MS4s’ concerns in implementing the first permits: “local governments argued that in many instances, biological and physical monitoring results are better indicators of small stream health.” Maryland’s NPDES Municipal Stormwater Monitoring § 4.1; MDE, Basis for Final Determination to Issue Anne Arundel County’s NPDES MS4 Permit (2013) (emphasis added). MDE acknowledged the local jurisdictions’ concern that “water ¹⁴⁶ chemistry data are of little value” in discerning the effects of stormwater in local streams when there is no assessment of stormwater as to “physical stream habitat and biological health.” MDE, Maryland’s NPDES Municipal Stormwater Monitoring at § 4.1. Nevertheless, MDE explained in its analysis that “[t]here is still a need to continue water chemistry work for the far field objectives of nutrient reduction in the Bay and accurate loading estimates for NPDES.” Id. at § 4.2.

In explaining its decision to increase the number of monitored storm events, MDE stated that “more frequent sampling” would “improve pollutant load estimates.” MDE, Maryland’s NPDES Municipal Stormwater Monitoring at § 4.0. Moreover, in explaining why it pared the number of chemical monitoring sites, MDE stated that the addition of biological and physical monitoring required the agency to reapportion resources. Id. at § 4.3. To compensate for the reduced number of sites, MDE explained that: (1) “[a] major goal for future monitoring will be to integrate water chemistry, biology, and physical assessments in one suitable watershed” to ensure “all three techniques [are] analyzed holistically”; and that (2) “site selection will be orchestrated at the state level” to “maintain an adequate number of residential, commercial, and industrial sites for State water chemistry needs.” MDE, Maryland’s NPDES Municipal Stormwater Monitoring at § 4.3. 63 MDE’s decision-making illustrates that, as the Policy Memo stated, the permitting authority and permittees worked together to “determine if storm water monitoring efforts [were] appropriate and useful.” 61 Fed.Reg. at 41,699.

MDE did not rubber stamp the jurisdictions’ proposal but affirmed that “[t]here is still a need to continue water chemistry work for the far field objectives of nutrient reduction in the Bay and ¹⁴⁷ accurate loading estimates for NPDES.” MDE, Maryland’s NPDES Municipal Stormwater Monitoring at § 4.2. 64 Moreover, biological and physical monitoring are within the scope of the EPA’s suggestions for alternative techniques, that is, “monitoring techniques other than end-of-the pipe chemical-specific monitoring, including habitat assessments, bioassessments, and/or other biological methods.” 61 Fed.Reg. at 41.699 (emphasis added). By increasing the number of monitored storm events, 65 MDE intended for the Counties to acquire more representative data about pollutant loads. Compare 61 Fed.Reg. at 41.699 (The EPA encouraged permit writers “to determine if ¹⁴⁸ storm water monitoring efforts are appropriate and useful.”), with 40 C.F.R. § 122.62 (d)(2)(iii)(D) (The initial application called for “[a] proposed monitoring program for representative data collection.”). 66 The Water Groups insist on interpreting representative in terms of the boundaries of the political subdivisions. In its analysis of the Counties’ initial monitoring experiences, MDE explained that site selection was a challenging undertaking, and one that it had to interpret flexibly so that the Counties could reasonably carry out their monitoring duties.

That is, “monitoring only five sites did not allow any single jurisdiction to monitor all possible land uses.” Maryland’s NPDES Municipal Stormwater Monitoring at § 2.3. From MDE’s perspective, however, the state-wide program to improve water quality was maintained by, “in the aggregate,” monitoring “the continuum from low density residential to heavy industrial sites.” Id. (emphasis added). The watersheds MDE has approved extend from low-density residential land use to commercial use and light industrial use. 67 Thus, we conclude ¹⁴⁹ MDE has approved of monitoring locations that adequately represent a continuum of land uses for the agency and the Counties to collect data.

See MDE, Maryland’s NPDES Municipal Stormwater Monitoring at § 4.3 (“[S]ite selection will be orchestrated at the state level” to “maintain an adequate number of residential, commercial, and industrial sites for State water chemistry needs.”) (emphasis added). MDE amended the program with the CWA’s objective at the forefront of its mind: “restoring] and maintaining] the chemical, physical, and biological integrity of the Nation’s waters.” Id.; see 33 U.S.C. § 1251 (a). In accordance with the Policy Memo, the agency’s changes and reasoning strike us as thorough efforts to make the monitoring program “appropriate and useful.” 61 Fed.Reg. at 41,699. We conclude that MDE acted within the discretion the EPA accorded it in the Policy Memo.

Thus, we also conclude that MDE has committed no legal errors pertaining to 40 C.F.R. § 122.26 (d)(2)(iii)(D). For many of the same reasons, we believe that the Permits contain monitoring “sufficient to yield data which are representative of the monitored activity.” 40 C.F.R. § 122.48 (b). The Permits require monitoring of many pollutants. Part IV.F.1.a-d. 68 Following the Counties’ first experience imple ¹⁵⁰ meriting MS4 permits, MDE increased the frequency of monitoring to “improve pollutant load estimates.” MDE, Maryland’s NPDES Municipal Stormwater Monitoring § 4.0.

The agency also decided to orchestrate site selection at the state level to “maintain an adequate number of residential, commercial, and industrial sites for State water chemistry needs.” Id. § 4.3. By taking control of the selection of monitored sites, MDE could better ensure monitoring of a continuum of activities (industrial, residential, and commercial). Additionally, the Counties will have to explain their monitoring process to ensure that the agency can evaluate whether they have complied with MDE’s requirements. Part V.A.l.b.

(The Counties shall submit annual reports and include “[a] narrative summary describing the results and analyses of data, including monitoring data that is accumulated throughout the reporting year.”). 69 ¹⁵¹ MDE also requires the Counties to assess WLAs, which represent pollutant loads. Part IV.E.2.b.iii. The Counties must record continuous flow measurements to obtain data to estimate pollutant reductions. Part IV.F.l.aiv.

Because the Counties are to achieve WLAs over time through restoration activities, MDE requires the Counties to conduct such monitoring “where the cumulative effects of watershed restoration activities can be assessed.” Part IV.F.l (emphasis added). MDE thus structured the Permits and imposed requirements on the Counties to ensure that they will monitor the discharge of pollutants to yield representative data. We conclude that the Permits comply with 40 C.F.R. § 122.48 (b). E. Permit Limitations 40 C.F.R. § 122.44 (i) requires monitoring “to assure compliance with permit limitations.” We begin by noting the two applicable limitations: (1) controls to reduce the discharge of pollutants 70 and (2) the restoration of 20% of impervious surface area not restored to the MEP.

See Part III.D; Part III.E.2. 71 ¹⁵² 1. Reduction of Pollutant Discharge The Permits include two important components that inform our analysis of whether MDE’s monitoring program ensures reductions in the discharge of pollutants: (a) the monitoring and modeling scheme and (b) adaptive management. As we explain, our initial discussion of monitoring and modeling is not sufficient to measure MDE’s compliance with 40 C.F.R. § 122.44 (i). Rather, it is a necessary component of what we consider a two-part scheme MDE designed to ensure the Counties would implement controls to reduce stormwater discharges. a.

Monitoring and Modeling MDE requires the Counties to implement a stormwater management program (“SWMP”), which consists of BMPs. Part IV.D.1.a. The Counties must integrate SWMPs with other permit requirements. Part IV.D. Aside from their role in the 20% restoration requirement, BMPs are most critical in the restoration plans that the Counties must implement to achieve progress in meeting WLAs.

Part IV.E.2, 4; see swpra note 17 (The WLAs are derived from the EPA approved TMDLs for each Permit.). To achieve progress in meeting WLAs, the Counties must first set a baseline for stormwater pollutant loads. MDE, Guidance, at 3. 72 To set this baseline, the Counties must apply TMDL pollutant loading rates to urban land use data. Id.

An MDE-approved model, such as MAST, will automatically calculate the cost and load estimates for Counties to inform their decision-making. See About MAST, Maryland Assessment Scenario Tool, available at http://www.mastonline.org/About. aspx [https://perma.cc/7NRV-QNXX]. The Counties can compare different scenarios to determine what BMPs they can ¹⁵³ implement to reduce pollution. Id.; see also MDE, Guidance, at 1 (“Local governments can weigh the cost associated with implementing different practices and choose the most efficient option for meeting pollutant load reductions.”).

The Counties select and begin implementing BMPs in accordance with their restoration plans. Part IV.E.2. Then they can monitor the effectiveness of the activities, Part IV.F.l.aiv, and use the monitoring information to determine if the activities have helped make progress in achieving WLAs, Part IV.E.2, 4. But the efficiencies for BMPs that the Counties select are based in part on estimates.

See, e.g., MDE, Guidance, at 10 (Table 4). 73 As our discussion of modeling illustrates, a model predicts the reductions a BMP can achieve. 74 But it does not tell the Counties exactly what the BMP will do in terms of pollution reduction. Indeed, many variables impact this process. 75 See Chesapeake Bay Program, Phase 5.3 Community Watershed Model, § 6.1.2, at 6-4 (“Uncertainty in estimates of ¹⁵⁴ BMP effectiveness is due to factors including (1) variability in precipitation, hydrology, soils, and geology; (2) variable performance of land management practices; [and] (3) lag time between implementing a practice and full performance and observed water quality benefits — ”); see also id. at § 6.2.2, at 6-8 (“Virtually all research data are generated under controlled management conditions” that, among other things, are designed to minimize hazards; thus, “the research estimates are more representative of a best-case scenario.”). But this should not raise a red flag about the legal correctness of the Permits. Indeed, the inherent shortcoming in estimating BMP effectiveness explains MDE’s decision to require the Counties to adapt their practices based on improved knowledge over time.

See infra Discussion: Adaptive Management. Rather, because the Counties must implement BMPs to make progress in achieving WLAs, estimation and prediction are necessary evils in this context. Thus, high quality assessments of BMPs are particularly important in the Permits so that MDE and the Counties can understand which practices are most effective to meet the many WLAs incorporated into the Permits. See infra note 17.

As we have discussed, MDE requires the Counties to assess BMPs through a focused monitoring approach. Part IV. F.1.a.iv. In the Basis for Final Determination, MDE explained that pollution reduction research supported the use of focused monitoring to better understand and, thus, improve BMPs.

MDE, Basis for Final Determination to Issue Anne Arundel County’s NPDES MS4 Permit