2Introductory Comments Several U.S. Federal agencies conduct national-scale periodic surveys to monitor status & trends of natural resourcesMost are conducted by U.S. Department of Agriculture (USDA) or Department of Interior (DOI)The setting: Current vs. Mid-1990’s vs. EarlierWill focus mostly on FIA & NRIQuick overview of programsHistorical endeavorsFt. Collins project (early 1980’s); Lund (1986); Leech (1998)“Realities of conducting natural resource surveys”
3Northern Oregon Demonstration Project – Overview Inter-agency demonstration project conducted in mid- 1990’s to examine feasibility of combining/integrating Federal environmental surveysFocused on 6-county area of Oregon that contains diversity of land cover & use, and ownershipsScientists from 6 agencies were responsible for funding, design, implementation, management, analysis [USFS, NRCS, NASS, USGS/NBS, BLM, EPA]
5Northern Oregon Demonstration Project – Introduction Support from Under Secretary’s office, Federal Geographic Data Committee (FGDC), and White House (CEQ) – but “hands off” approachThe project goal was to study broad topic of integrating natural resource surveys – but specific focus was on NRI, FIA, FHM, and NFS survey proceduresGoebel, Schreuder, House, Geissler, Olsen, and Williams (1998); House et al (1998)Many issues and concerns were identified, but project focused on 7 objectives
6Northern Oregon Demonstration Project – Objectives Ascertain if sampling frames give proper coverageDetermine “best” frame; investigate statistical & operational difficulties of constructing joint data baseExplain discrepancies in forest & range (area) estimates
7Northern Oregon Demonstration Project – Objectives Investigate collecting common information on common samples with joint FIA/NRI data collection teamsExplore data collection methodology for vegetation & soil attributes in integrated survey contextDetermine whether sampling for animal abundance can be included in survey designAnalyze measurement errors associated with collection of different variables [most important for new protocols]
8Northern Oregon Demonstration Project – Data Collection Design & Methods Data collection portion conducted in 3 phasesIncluded selection of important existing measurements from NRI, FIA, FHM, and NFS Region 6 surveysAlso included several experimental variables associated with soil quality, range and forest health, wildlife habitat, and animal relative abundance
9Data Collection – Phase I Carried out in office by experienced USFS, BLM, and NRCS personnelUsed aerial photos, GIS data layers, hard-copy ancillary materialsSample consisted of 613 sample points: 337 FIA/NFS sites and 276 from NRIsamples selected independently from two complete frames, soused straight-forward multiple-frame estimation proceduresData elements: several cover & use, classifications, evidence of disturbance, soils, site characteristics ownership category, geographic delineations (e.g., HU)
10Data Collection – Phase II Carried out by joint 2- and 3-person field crewsUSFS personnel were FIA inventory specialistsNRCS: soil scientists, soil conservationists, & rangeconservationists [with some NRI experience]Sample consisted of 91 sample points selected from the 613 Phase I sample sites [unable to sample 13 sites]Data elements: site characteristics; veg. structure; ground cover; herbaceous veg. species freq.; shrub canopy cover; shrub density; tree tallies; woody debris; soil characteristicsSoil samples collected & analyzed at soil laboratoryAll variables collected for each sample but various protocols used to obtain different measurements
12Data Collection – Phase III Carried out by specialized 3-person USGS field crew [National Biological Survey staff]Sample consisted of 14 Phase II sample sites occurring on particular portions of 3 national forestsVarious protocols used to observe diurnal breeding birds, amphibians, ground insects, and flying insectsEach site visited 3 times within 5-week period
13Measurement Repeatability Study (Data Collection) Each Phase II sample site was visited by 2 different crewsSubplots 1 & 2 sampled by both crews; only one crew sampled subplots 3 & 4Plot data collected independently by the 2 crewsVisits by the 2 crews made at same timeOperational efficiencyLimited accessibility to private propertyEnsured that measurements made at same locations
14Some of the Lessons Learned Agencies can work together; have complementary skillsUniform land classification is achievableMany basic inventory needs can be met with the same protocolsSensitivity of access to private landsEfficiencies of doing things only once is achievablePlant identification to species level = large workloadMust have mobile GPS units and CASI (Computer Assisted Survey Instrument) – more than just a data recorderDeveloped an “Integrated Inventory Vision”
15Forest and rangeland estimates (in ha Forest and rangeland estimates (in ha.) using USFS and NRCS definitionsForest Land RangelandCrown USFS NRCS USFS NRCSLand Class Cover % Estimate Estimate Estimate EstimateTimberland , ,517, ,972OakWoodland , ,036, ,358UnclassifiedWoodland, ,361JuniperWoodland , ,403, ,912Chaparral , ,036Desert Shrub , ,548Grass/Herbaceous , ,820Total (Phase I) , , , ,27245% % % %Total – Regression , , , ,913Estimator % % % %
16Repeatability of Selected Measurements Correlation Measurement error as(r) % of plot varianceAverage # of plantspecies per plot %Average DBH %Total basal area %Number of species %Number of trees %% of total shrubs as seedlings %% of total shrubs as mature %Total count, shrubs %
17Interagency Inventory & Monitoring Initiative (AIIMI) Follow-up to Northern Oregon Demonstration ProjectStudy area = Minnesota; initiated in 1999Further explored feasibility and limitations of integration (of FIA and NRI)Featured assimilation & use of data rather than new data collectionFurther examined differences in focus & design of inventories when combining data in a common frameworkCollaborators: Minnesota DNR; USFS; NRCSAlso USGS EROS Data Center for one projectNRCS Statistician co-located with FIA in St. PaulCzaplewski et al (2002); Rack et al (2002)
18AIIMI - Products GIS Test Data Base GIS test-bed provided a statewide integrated coverage of FIA, FHM, NRI, and variety of other (ancillary) spatial dataHuge task; quite valuableAncillary data included: STATSGO soils data; Census data; Digital Elevation Model (DEM) data; Digital Raster Graphics (DRG) data; supplemental digital aerial photography; Landsat TM imagery; Digital Ortho Photo quads; wetlands and ecological zone mappingIntranet Application for Retrieving and Viewing Plot-level Imagery and GIS DataNavigational capabilities enable data collection and QA specialists to view plot locations in a landscape context
20DescriptionThis shaded polygon map shows the amount of forest land converted to developed land between 1982 and 1997 within each 8-digit hydrologic unit. The acres converted are presented in five categories based on the following divisions: 20,000 or more acres, 5,000 to 20,000 acres, 2,000 ro 5,000 acres, and less than 2,000 acres. A total of 10.3 million acres were converted. Developed land includes urban and built-up areas and rural transportation land. Metropolitan Statistical Areas are indicated by black squares. Areas with 95% or more Federal area are shaded gray.Cautions for this Product: This map does not show the total amount of forest land or developed land. Data are not collected on Federal land. Data are not available for Alaska or the Pacific Basin. Data for Puerto Rico and the U.S. Virgin Islands are shown by 6-digit hydrologic unit.Center point of Central City for each MSA with a population of 100,000 or more.
21AIIMI - Products (cont.) Comparison of FIA and NRI EstimatesInvestigated land cover/use classification and area estimates to discover types and reasons for similarities and differences in estimatesMapping Changes in Land Cover/UseBased upon both FIA & NRI plot dataGeospatial representation of changeProvides insight and perspectives not available through commonly reported summary statistics
22AIIMI - Products (cont.) Image-based detection of land cover changeUsed integrated set of FIA and NRI data for 10-county area as training data for classificationLandsat classification utilizing NRI and FIA plot dataConducted in cooperation with USGS Data CenterTo determine if FIA and NRI data would help in development of National Land Cover Data (NLCD) mapping
23AIIMI - Discussion; Findings GIS DataIt takes considerable work to “align” geospatial dataMostly manual work rather than automaticDiffering standards, scales, etcCover and Use DataClassification systems vary between programsNRI and FIA oriented toward use; satellite data – coverFor plots giving heterogeneous signatures – difficult to correlate satellite and survey plot data
24AIIMI - Discussion; Findings (cont.) Maps – Geospatial Displays of DataVery useful in supplementing area statistics [for example, where are the losses of forest land to urban development]Requires spatial and temporal consistencyAnnual InventoriesBoth FIA and NRI migrated to Annual Inventory system during the period that AIIMI was being conductedBoth surveys being “annual” should help collaborative effortsBut both programs were too pre-occupied with implementation (including funding issues) to seriously investigate integration
25AIIMI - Suggestions Use GIS to develop common “Universe of Interest” NRI & FIA should have same Total Surface Area & Census WaterDevelop common “cover” classification systemWould allow USDA to have “common reporting system”But also – FIA and NRI need to keep their current/historical systems [needed for Agency programs & have huge investment]Soils DataAdd NRCS soils data base information to FIA, geospatially [would have characteristics and interpretations for each sample site]FIA would then supply plot information to NRCS to enrich the soils data bases [productivity; biomass]
26AIIMI - Suggestions Further linkage of FIA and NRI data StatisticalgeospatialSurvey IntegrationCzaplewski et al (2002)]Limited budgets; Accountability; OMBDo NOT start from scratchUtilize strengths of each systemNRI: land use change; soil; cost/ plot; site condition (general)FIA: volume; veg. composition change; site condition (specific)
27.FIA/NRI Integration – should take advantage of each program’s strengths & not start from scratch
28Other Inter-Agency Efforts Status and Trends of WetlandsAssessment of RangelandsNorth American Carbon ProjectAgricultural StatisticsResource Inventory & Monitoring, Focus Area Work Group (FAWG), NASA/USDANational Land Cover Characterization, NLCD 2001
29Status & Trends of Wetlands National estimates produced through 2 separate naturalresource surveys [both with legislative mandates]Status & Trends – USFWS, Dept. of InteriorNRI – NRCS, USDAConsiderable pressure during the 1990’s to develop a single report by year [Clean Water Act]Currently not possible to produce statistically reliable results by combining USFWS and NRI data [Dahl (2000)]AccomplishmentsJoint press conference Jan. 2001, Secretaries of Interior & AgricultureStatistics on trend (Quantities & types of loss) are “consistent” due to field work by USFWS & NRCS, and subsequent report modifications
30Assessment of Rangelands National Research Council (1994)Called for development & utilization of new methods to classify, inventory, and monitor rangelandPlaced emphasis on rangeland healthsCooperative work during 1995 – 2002 to develop field protocols that attempt to address Council’s callNRCS, ARS, BLM, & USGS have been most active, with limited participation by USFSWhat about “Criteria & Indicators for Sustainable Rangeland” [Sustainable Rangeland Roundtable]?Protocols meant to help detect long-term changes in conditions & to monitor short-term impacts
31Development of Rangeland Protocols Limited trial studies started in 1996 in 2 regionsBLM conducted field test in Colorado, 1997 & 1998Limited field test conducted on private lands in 7 states in 1999Collected valuable cost/time dataCurrent protocols include combination of quantitative and qualitative measurementsNRCS utilizing these as part of NRI for 2003 – 2005NRCS expects that a subset of these will be “permanent”Research activities (with ARS) – reduce replications; incorporate remote sensing; make 100%quantitative
32Current Rangeland Protocols Ecological site information; soils; landscapeLine point transects for cover compositionLine intersect transects for basal & canopy coverCover density & height [wildlife habitat]Disturbance indicators; conservation practices & treatment needsNoxious weeds & invasive/alien plantsSoil stability testSpecies composition by weightRangeland Health
33North American Carbon Project Need complete accounting for carbonInvolves many Agencies, Universities, etc.Science-based approachFor both domestic and international reportingNeed to reconcile models [& calibrate & improve]“Top down” approach [Atmospheric scientists]“Bottom up” approach [Agricultural & forestry scientists]
34Soil carbon in forested lands of the North Central region
35OpportunityAs part of the North American Carbon Project, there appears to be a need to build a comprehensive FIA/NRI Data BaseReconcile FIA & NRI data for use in C models & elsewhereOne “proposal” is to create geospatial (tesellated) data base with land use, land management, land use history, soils [maybe something equivalent to 10-km. grid ??]Would include measures of “uncertainty”Would need protection of confidentialityShould also investigate incorporation of NASS crop maps, MODIS data, and ???
37Agricultural Statistics NASS & NRCS currently cooperating on several survey activitiesReconciliation of NRI and Census of Agriculture acreage figures – showing how to properly align categoriesConservation Effects Assessment Project (NRI-CEAP), where NASS conducting 0n-farm interviews for NRI sample sites; Farm Services Agency (FSA) also cooperatingInvestigating integration of Agricultural Resource Management Survey (ARMS) & NRI-CEAP, collaboratively with Economic Research Service (ERS)NRI needs NRI-CEAP type data on an annual basis for many uses (including C modeling) – part of Continuous NRI concept introduced in 1998NASS crop maps
38Resource Inventory and Monitoring, Focus Area Work Group (FAWG) One of 8 focus areas identified by NASA and USDA in May 2003 MOUObjective is to identify projects for collaborative development to enable USDA operating units to incorporate NASA earth observations, modeling, and systems engineering capabilitiesNRI and FIA serving as co-chair
39National Land Cover Characterization (NLCD), 2001 Land cover data base being developed by region/zoneCooperative mapping effort of Multi-Resolution Land Characteristics (MRLC) 2001 consortiumUSGS EROS Data Center collaborating with EPA, USFS, NOAA, NASA, NPS, USFWS, BLM, NRCS (NASS?)Utilizes Landsat TM data from 3 time periods, plus ancillary data from Digital Elevation Model (DEM)Zone 41 (much of Minnesota) – developed as part of AIIMIProduces “objective” data layers for each time periodDecision tree approach – rules developed to transform objective data into themes [cover; imperviousness; trees]
40The Realities of Conducting Natural Resource Surveys – Lessons Learned Who pays the bills? What pays the bills?What is expected of your survey program?When do we get “burned”?How do we maintain “credibility” with Policy Makers, other scientists, the public? Perception is almost everything. Cooperating with an independent entity like Iowa State University is good business & good science!!“Keeping NRI going” is a large challenge. Therefore, inter-agency is even greater challenge?
41The Realities of Conducting Natural Resource Surveys – Lessons Learned Who pays the bills? What pays the bills?“MONITORING” – conducting a longitudinal survey properly for natural resources rather than for people issues [health; economics] – are the scientific and operational challenges fully realizedNew (& great) technologies come along that affect your “favorite reporting indicator”, like soil erosion for NRI. What do you do?Are you sampling farms or fields or forests or trees? What happens with departures and new arrivals into your universe of interest?
42The Realities of Conducting Natural Resource Surveys – Lessons Learned Who pays the bills? What pays the bills?“MONITORING”Indicators [condensing complicated science into useful factoids] – collect the “most basic factors” and not the “Indicator” itselfOMB/USDA Quality of Information standardsRealistic – must use Computer Assisted Survey Instruments & modern supporting systemsMake sure that you can deliver – No excuses!