u0198116/Aug_9_NOAA/Complet… · Web vie

E. Crosman: Improving Long-term Records of High-Resolution Satellite-derived Lake Surface Temperature for Global and Regional Climate Studies

Research Proposal Submitted to: NOAA CPONOAA (R/CP1), SSMC3, Room 12734, Silver Spring, MD 20910

Improving Long-term Records of High-Resolution Satellite-derived Lake Surface Temperature for Global and Regional Climate Studies

Federal Funding Opportunity Number: NOAA-OAR-CPO-2018-2005133Competition: CPO – Ocean Observing and Monitoring Program (OOM)

Erik CrosmanResearch Assistant Professor

Department of Atmospheric Sciences135 S 1460 E, Room 819 WBBSalt Lake City, UT 84112-0102

University of UtahPhone: 505 570-0552; Fax: 801 581-4262

Email: [email protected]

Institutional Representative: Erica TrejoOffice of Sponsored Projects, University of Utah

1471 E Federal WaySalt Lake City, UT 84102

801 581-6232; 801-581-3007 (fax)[email protected]

Funds requested: TotalYear 1 135,617; Year 2 $132,807; Total $268,424Proposed project period 1 July 2018-30 June 2020

Funds requested for University of Utah: Year 1 135,617; Year 2 $132,807; Total $268,424

1


Improving Long-term Records of High-Resolution Satellite-derived Lake Surface Temperature for Global and Regional Climate Studies

Federal Funding Opportunity Number: NOAA-OAR-CPO-2018-2005133Competition: CPO- Ocean Observing and Monitoring Program (OOM)

PI: Erik Crosman, University of Utah

Abstract (not to exceed one page)Lakes are a critical component of the earth system and are known as “sentinels” of climate change. Lake surface water temperature (LSWT) is an important parameter for quantifying and modeling regional responses to climate change and LSWT has been shown to have impacts on regional climate. However, considerable uncertainty in the trends calculated from existing satellite-derived LSWT exists, and global climate models do not resolve many lakes and/or do not use appropriate surface temperature forcing data sets to prescribe LSWT in retrospective climate simulations. The proposed work will utilize satellite-derived lake temperature datasets from two state-of-the-art global climate SST products (NOAA Pathfinder V5.3 and NASA MODIS LST V) to produce new quality-controlled LSWT climatological and trend data sets with uncertainty estimates between 1985-preent for use by global climate change modeling and observational studies. The summary of the work to be completed is threefold:

1) Produce a new, improved global lake temperature mean climatology data set as well as a long time series trend dataset for hundreds of lakes for use in climate change observational and numerical modeling studies derived from the NOAA Pathfinder V5.3 and NASA MODIS LST data sets, with a focus on both cold-season (when clouds limit number of available observations) and smaller lakes (5-30 km in diameter), as daily time series of these products are currently unavailable.

2) Determine the ability of satellite-derived lake temperature datasets to evaluate global trends in lake surface temperature between 1985-present. Rigorous in situ validation of satellite retrievals will be conducted for numerous lakes over long periods of record, which has not been previously conducted.

3) Apply improved cloud masking and other statistical quality-control algorithms to produce an improved lake temperature time series, and quantify the potential impact of errors and uncertainties associated with cloud contamination and temporal gaps (long cloudy periods) on satellite-derived lake temperature trends.

The proposed research targets the following areas in the Ocean Observing and Monitoring FY18 program call for proposals: (1) Development of data sets for the climate research community and (2) Projects that develop or improve datasets suitable for periodically updated assessments or monitoring products for weather and climate extremes and impacts on water resources. This work will provide the basis for improving existing as well as providing new climate data for satellite-derived LSWT data sets, and will produce targeted research results that will improve uncertainty in the error characteristics of satellite-derived LSWT retrievals for use in long-term monitoring and climate modeling. In terms of NOAA’s long term climate goals, this proposal will advance climate intelligence and resilience through providing a data set useful for addressing (1) weather and climate extremes, (2) climate impacts on water resources, and the (3) sustainability of marine ecosystems.

2


Results from Prior Research

1. Multi-sensor Improved Sea Surface Temperatures (MISST2): 2013-2017 [PI E. Crosman; Co-PI J. Horel]

This project, which ends in December 2017, is being supported by the NASA Integrated Ocean Observing System (IOOS) program. The goal of this study was to provide recommendations for improving lake temperatures for input into numerical weather prediction models, and to review the error sources and uncertainties associated with lake temperature retrievals, as well as the various split-window and other algorithms that have been developed for satellite-derived lake temperature over the past few decades. As part of this study, an evaluation of the NASA Multi-scale Ultra-high Resolution (MUR) analysis of lake surface temperature for several lakes was conducted, with results recently published in Crosman et al. (2017a). Statistical and other quality-control algorithms specifically designed for improving lake temperature retrievals from satellite have been developed at the onset of this study and were presented by Grim et al. (2013). An extensive review paper on lake temperature (Crosman et al. 2017b) has been submitted.

2. NMP Data Hub. 2017-2018 [PI E. Crosman]As part of this program, the University of Utah provides a wide range of weather data under the National Mesonet Program for use by the National Weather Service, public, and private sectors, and for input into the National Centers for Environmental Prediction Meteorological Assimilation Data Ingest System (MADIS) for initializing a suite of numerical weather models. As part of this project, a network of weather stations are maintained in northern Utah as part of this project, including 14 surface stations measuring wind, temperature, humidity, solar radiation, atmospheric pressure, precipitation. Several wind sodars and backscatter ceilometers for profiling the boundary-layer are also operated on a routine basis as part of this project. The data for this project are all available via a state-of-the art API: https://synopticlabs.org/api/ . This includes a number of in situ buoy measurements of lake temperature that will be utilized in the proposed study.

3. Improve Air Quality Modeling for the Wasatch Front & Cache Valley Winter Air Pollution Episodes: 2015-2016 [PI E. Crosman]

The Utah DAQ supported a modeling study to develop appropriate modeling strategies to simulate the meteorological conditions associated with poor winter air quality episodes in Utah basins. This work builds on the prior modeling work for similar conditions in the Uintah Basin (Neemann et al. 2014) as well as other wintertime modeling studies for the Salt Lake region (Lareau and Horel 2015a,b). The simulated meteorological conditions are sensitive to the specification of the land and lake surface state, e.g., areal extent and temperature of the Great Salt Lake and surrounding snow cover. The results of this research were published by Foster et al. (2017).

4. 2015 Great Salt Lake Summer Ozone Study: 2014-2016 [PI J. Horel; Co-PI E. Crosman]The Utah Division of Air Quality (DAQ) supported a field study in 2015 to improve understanding of the temporal and spatial distribution of ozone in the vicinity of the metropolitan regions of northern Utah (Horel et al. 2016). During the 2015 summer months, DAQ staff and researchers from the University of Utah, Utah State University, and Weber State University

3

https://synopticlabs.org/api/


deployed ozone sensors near the Great Salt Lake at fixed sites as well as on a UTA TRAX light rail car, vehicles, UAV’s, and tethered balloons. An ozone sensor onboard the KSL traffic helicopter (Crosman et al. 2017) provided critical information on ozone concentration aloft, particularly along major traffic corridors during the late afternoon when ozone levels reach their peak. The temperature of the Great Salt Lake derived from satellite (Blaylock et al. 2017) was found to be a contributing factor in the ozone pollution levels associated with the lake breeze front. The data from this project are publically accessible at the following website:http://meso2.chpc.utah.edu/aq/cgi-bin/mobile_data.cgi

4

http://meso2.chpc.utah.edu/aq/cgi-bin/mobile_data.cgi


Statement of WorkI. Overview1. Problem StatementOur proposed research focuses on providing a new long-term and quality-controlled satellite-derived lake surface temperature climatological dataset between 1985-present for use by the climate sciences for a wide range of applications. Only sparse in situ observations of lake temperatures exist, primarily in European and North American lakes. Consequently, long time series of satellite-derived lake temperature with global coverage is of high potential value for a wide ranges of geophysical applications and climate studies. Satellite-derived lake surface water temperature (LSWT) estimates are generally more uncertain and unavailable to the scientific and general public than oceanic sea surface temperature (SST) retrievals (Oesch et al. 2008; Hulley et al. 2011; Fiedler et al. 2014). In addition, no comprehensive global quality-controlled and extensively validated climatological global data sets of daily LSWT for small lakes (<25 km in diameter) exists. Evaluation of several recent climatological studies of LSWT for (mostly) large lakes (Hook et al. 2012; Layden et al. 2015; O’Reilly et al. 2015) indicates uncertainty and in some cases opposite climate trends deduced from the various LSWT climate studies, suggesting that documented problems with cloud contamination (Oesch et al. 2008; Politi et al. 2012) and temporal gap errors (Crosman et al. 2017a) may be introducing uncertainty in the existing analyses of global LSWT trends on these medium-sized to large lakes, These differences indicate the need to evaluate these products globally and to apply quality-control measures to mitigate the potential impacts of cloud contamination and temporal gap errors on satellite-derived LSWT trend analysis and LSWT climatological data sets.

2. Scientific ObjectivesIn this study, we will develop and implement cloud clearing and statistical quality control algorithms for use with satellite-derived lake temperature datasets from two state-of-the-art global climate SST products (NOAA Pathfinder V5.3 and NASA MODIS LST V) and then derive new data sets of quality-controlled LSWT climatological annual cycles in LSWT and trend daily time series with associated error and uncertainty estimates for use by global climate change modeling and observational studies. The relevancy of this data will only increase as regional climate modeling studies are run at increasingly high spatial resolution. The MODIS radiometer (2000-present) has both the high-resolution capable of resolving smaller lakes and now a sufficient period of record (2000-present) to resolve climate trends in “smaller” lakes between 4 and 20 km in diameter. These smaller lakes have not been rigorously studied or validated to our knowledge on a global scale in previous studies. In addition, the latest version of NOAA Pathfinder V5.3 (1981-2014) that was recently released includes a number of improvements that likely improve data availability over larger small lakes and over some medium-sized lakes that have not been previously analyzed (16-40 km in diameter). In this study we will also rigorously validate both the pre-processed and post-quality control LSWT datasets against in situ observations on a number of lakes. More rigorous validation over long time periods and over small lakes between in situ and satellite-derived LSWT is needed, as most lake validations were conducted over large lakes and temporally short periods of time (Crosman et al., 2017b).

5


3. Task Statement RelevancyWe will address two of the three of the research needs identified by the FY18 Ocean Observing and Monitoring Division call for “High-quality data sets for enhancing predictions and informing stakeholders: (1) Development of data sets for the climate research community and (2) Projects that develop or improve datasets suitable for periodically updated assessments or monitoring products for weather and climate extremes and impacts on water resources. Restated in the context of our proposed research, this research will (1) Develop improved LSWT data sets for the climate research and modeling community, and (2) develop and improve by rigorously validating and quality-controlling available satellite-derived LSWT records to assess the impacts of changes in global LSWT for climate, weather, and water resources. The time series and climatology of LSWT produced by this study will directly address NOAA’s long term climate goals to advance climate intelligence and resilience through providing a data set useful for addressing (1) weather and climate extremes, (2) climate impacts on water resources, and the (3) sustainability of marine ecosystems.

4. BenefitsThe PI has a demonstrated record in evaluating lake surface temperature analyses and trends from satellite that provide benefit to the research and operational communities (Crosman and Horel 2009; Grim et al. 2013; Crosman et al. 2017a, b). Our research will provide the largest benefits to the general public and scientific community by improving both the accessibility and quality of available lake surface temperature data, which provides valuable information on lake state for a wide range of geophysical applications that are sensitive to global climate and climate change (Fig. 1). Currently, much of the climate and limnological community does not view or utilize satellite-derived lake temperature as it is not an easily-accessible product and documentation does not exist (many users do not realize that oceanic SST data sets are also generated over inland waters), unlike the highly accessible nature of numerous SST products for the wide oceanographic community. Deliverables will be focused on providing new derived data sets of quality-controlled LSWT climatological and trend daily time series for the large community that would benefit from the data. The deliverables will also include at least two peer-reviewed research articles and webinars

6

Figure 1. Key application areas of satellite-derived lake surface temperature


and presentations in national and regional venues oriented to the climatological, limnological, biological, and numerical weather prediction and climate communities. In addition, the outcomes of our research will help develop recommendations for future enhanced reprocessing with lake-specific split-window algorithms of the entire period of satellite record of temperature of inland waters in terms of appropriate quality flags, cloud masks, and statistical tests to filter residual cloud contamination by thin cirrus.

II. Technical and Scientific BackgroundLakes worldwide provide many far-reaching benefits to society, including drinking and agricultural water, fishery habitat, recreational opportunities, transportation routes, and hydroelectric energy (Stenderra et al. 2012; Dornhoffer and Oppelt 2016) (Fig. 1). As an integral component of the earth system, lakes have been found to be “sentinels” of climate change (Adrian et al., 2009; MacKay et al., 2009; Williamson et al., 2009; Castendyk et al., 2016). Approximately 117 million lakes worldwide cover 3.7% of the non-glaciated planetary land surface area and comprise a total volume of ~200,000 km3, with over 17,000 lakes worldwide with surface areas greater than 10 km2 (Verpoorter et al., 2014; Cael et al., 2017). Satellite remote sensing of lakes can provide valuable information on lake water transparency, biota, hydrology, temperature, and ice phenology (Dornhoffer and Oppelt 2016). Lake surface temperature is an important parameter for understanding and modeling the biology, hydrology, weather and climate of lacustrine and adjoining terrestrial environments (e.g., Dutra et al. 2010; Kraemer et al. 2016; Reavie et al. 2016; Javaheri et al. 2016; Mason et al. 2016)(Fig. 1). Lake surface water temperature (LSWT) retrievals from satellite thermal infrared (TIR) sensors onboard numerous satellite platforms provide a spatially comprehensive dataset for lakes in the absence of clouds, and are the primary means used to obtain LSWT over thousands of lakes worldwide

7

Figure 2. High-resolution analysis at 1 km of Lake Temperature by the NASA MUR analysis (see Crosman et al, 2017a, Chin et al. 2017) in four global sub-regions with high density of small to medium-sized lakes. Imagery courtesy of NASA State of The Ocean https://podaac-tools.jpl.nasa.gov/soto/.

https://podaac-tools.jpl.nasa.gov/soto/

https://podaac-tools.jpl.nasa.gov/soto/


where in situ temperature measurements are unavailable (e.g., Fig. 2). Remote sensing of LSWT began in the 1980’s and has increased steadily in subsequent decades. However, satellite-derived LSWT estimates are generally more uncertain and unavailable to the scientific and public than oceanic sea surface temperature (SST) retrievals due to an increased difficulty in correcting for continental atmospheric air masses, lake elevation, cloud contamination, and shoreline effects (Hulley et al. 2011). In addition, gaps in the available satellite images due to clouds, the small size of many lakes (too small to be measured by many satellite infrared sensors) and relatively sparse in situ observations make it difficult to obtain imagery during cloudy and cool seasons in the mid-latitudes, and to validate these satellite-derived LSWT and to provide calibrated long-term spatially- and temporally-consistent LSWT analyses.

LSWT is also a critical input variable for numerical weather, climate, and hydrological models (e.g., Dutra et al. 2010; Balsamo et al. 2012; Kheyrollah Pour et al. 2014; Javaheri et al. 2016). While extensive climatological data sets and analyses of satellite-derived sea surface temperature (SST) are available and used in a wide range of applications, limited satellite-derived LSWT climatological data sets exist, and the quality of these retrievals is impacted by difficult to remove thin cloud cover, as well as a number of other challenges including lake elevation, variations in atmospheric profiles of temperature and moisture, and dust and aerosol (Crosman and Horel 2009; Hulley et al. 2011; MaCallum and Merchant 2012). No climatological data set utilizing high-resolution daily MODIS satellite data currently exists for prescribing the temperature trends of thousands of small lakes (5-25 km in diameter) worldwide, despite the known impacts of these water bodies on regional climate and weather. Only limited regional climatologies of small lakes have been conducted. The most recent ARC-Lake version 3 data set has been extended to include smaller lakes, reservoirs, and ephemeral lakes (http://www.laketemp.net/home/targets_phase3.php). However, this climatology is restricted by the ~weekly availability of satellite data from the Along-Track Scanning Radiometer (ATSR) imagery utilized in the ARC-Lake climatology. Only regional climatologies have been processed thus far with daily MODIS satellite data. For example, a recent study by Wan et al. (2017) derived a MODIS climatology between 2001 and 2014 for lake temperature for 374 lakes on the Tibetan plateau with areas greater then 10 km2, illustrating the feasibility of extending a similar approach globally as proposed for this study, while Riffler et al. (2015) derived a climatology for 25 larger lakes using local area coverage AVHRR data in lakes in the European Alps. The only global lake climatology currently available for ~200 of the (mostly) larger lakes in the world is the ARC-Lake climatology (MacCallum and Merchant 2012; Layden et al. 2015). Because of the frequent cloud cover in many mid-latitude regions, the lack of a ‘background’ climatological lake temperature product is a major limiting factor in the ability of weather and climate models to properly represent the surface state in regions such as Canada, the Tibetan Plateau, and the Rift Valley of Africa (Fig. 2). In this study, we propose to generate a high-resolution climatology for lakes which will directly address this problem. As discussed in the problem statement for this study, the limited climatological studies of LSWT that have been conducted for (mostly) large lakes (Hook et al. 2012; Layden et al. 2015; O’Reilly et al. 2015) have in some cases found variable rates of warming and cooling in the same lakes, as well as some lakes where opposite climate trends were deduced. This uncertainty suggests that careful analysis of the impacts of temporally and spatially variable cloud contamination and data gap sampling issues (Oesch et al. 2008; Politi et al. 2012; Crosman et al. 2017b) need to be investigated as these may be introducing uncertainty in the existing analyses of global LSWT trends on these medium-sized to large lakes, These differences indicate the

8

http://www.laketemp.net/home/targets_phase3.php


need to evaluate these products and apply quality-control measures and to quantify the potential impacts of cloud contamination and temporal gap errors on satellite-derived LSWT trend analysis. As part of this study, we will investigate these impacts, as well as provide the first trend analysis data for small lakes worldwide not included in the earlier studies. This study will provide quality-controlled lake temperature data to complement the lake state data from a subset of lakes being analyzed by the ongoing ecology-focused Globolakes project (http://www.globolakes.ac.uk/overview.html), which is investigating the chemical, physical and ecological condition of 991 lakes worldwide.

Most satellite LSWT retrieval algorithms were designed for ocean surfaces and validated and tuned to oceanic in situ buoy observations (Hulley et al. 2011; Fiedler et al. 2014). Consequently, the effects of variations in lake elevation, atmospheric profiles of temperature and water vapor, dust and smoke sources, and near-shore pixel contamination by adjacent land surfaces are not typically incorporated in the algorithms when they are applied over inland water bodies. Developing lake-specific algorithms for satellite-derived LSWT is an active area of ongoing research and several studies have developed improved techniques for LSWT retrievals (e.g., Hook et al. 20003, 2007; Hulley and Hook 2011; MaCallum and Merchant 2012; Layden et al. 2015; Riffler et al. 2015). However, the improvements gained by implementing lake-specific LSWT retrievals compared to utilizing oceanic algorithms are relatively modest in the limited lakes where validations have been conducted (Hulley and Hook 2011; MaCallum and Merchant 2012; Layden et al. 2015). The ARC-Lake project (MaCallum and Merchant 2012; Layden et al. 2015) is the only global LSWT that has utilized lake-specific split-window algorithms in reprocessing global lake temperatures to our knowledge, although Hulley and Hook used weather model data to compute split-window algorithms coefficients for a number of larger lakes worldwide. While reprocessing the entire SST data set for small to medium sized lakes with the MODIS and AVHRR raw radiance data combined with global atmospheric reanalysis data is a future goal, greater resources and time than available for this study would be required for this effort, as no data set currently exists for split-window algorithm coefficients for the numerous small lakes worldwide. In addition, the problems with cloud contamination and gap sampling errors appear to be much larger and more immediately problematic to the obtaining adequate climate trend analysis from satellite-derived SST (Crosman et al. 2017b). Thus, creating an annual climatology for smaller lakes worldwide, and an improved quality controlled data set for all lakes worldwide is the focus of this study. The currently publically-available (to our knowledge) LSWT climate data sets are listed in Table 1. The extensive validation of the LSWT alongside this effort will be critical to evaluate the improvements obtained by the quality-controlled climatology. Cloud contamination of satellite-derived imagery is a serious problem over lakes (Crosman and Horel 2009; Oesch et al. 2008; Politi et al. 2012 Fiedler et al. 2014). Improved cloud masking is also a need within the land surface remote sensing community (Li et al., 2013). Cirrus clouds are relatively common but difficult to screen for over land. Spatial homogeneity tests typically used for SST over oceans can incorrectly flag spatial variations in LSWT as resulting from cloud contamination, when in reality these spatial patterns exist as a function of the complex lake morphology. Cloud masking algorithms that flag elevated aerosol and dust are also needed. While a number of improved lake-specific cloud masking algorithms have been developed, they are not applied to most readily available satellite-derived LSWT products, and have not been tested on a wide range of lakes (Merchant et al. 2005; Hulley 2009; MacCallum and Merchant 2011; Fan et al. 2015). As seen in Fig. 3, visual removal of residual cirrus

9

http://www.globolakes.ac.uk/overview.html


contamination during a 2-week period in October 2015 over the Great Salt Lake, Utah resulted in up to several °C difference in the mean LSWT of the lake during this period. The existing state-of-the art AVHRR (CLAVRR) cloud masking algorithm was applied in both cases but was ineffective in removing thin cirrus from the retrievals. In addition to cloud contamination problems, some of the cloud masking algorithms used over the ocean are too stringent and incorrectly flag clear sky pixels as land in some data sets. The new Pathfinder V3.5 data set utilized in this study now includes the lowest ‘0’ quality level data, which will be analyzed to determine how many lakes are losing valuable data due to these problems when only higher quality level data is utilized. In this study we will leverage statistical cloud screening algorithms with other statistical QC procedures (e.g., Grim et al., 2013) that use expected climatological and spatial LSWT characteristics to constrain the limits on threshold and spatial homogeneity tests. Some currently applied cloud masking tests that are used in SST retrievals, such as spatial heterogeneity tests, may need to be relaxed or removed over some lakes due to the large spatial variations in LSWT, and this study will provide guidance that is not currently available with respect to these issues. In addition to extensive clearing of cloud contamination, we will also analyze case studies of the impacts of large dust storms and wildfires (e.g., persistent smoke in western U.S. and Canada in summer 2017) to determine the potential impact of these temporally variable sources of error on seasonal LSWT averages.

Table 1. Comparison of current publically-available LSWT data sets and the LSWT data set proposed to be developed for this study

Lake Surface Water Temp Data Set

ΔX Period of Record

Filtering for residual cloud contami-nation

Ability to Resolve smallLakes (< 20 km)

Rigorous validationand QC

Post-processed time series and derived climatology for NWP

1. NOAA Pathfinder V5.3

~4 km

1981-2014 X

MODIS SST/LST

~1 km

2000-present

X

ARC-Lake ~ 6 km

1991-2012 X

X

Sharma et al. (2015) summer lake temperatures

~ 4 km

1985-2009

Proposed LSWT data set in this study

1 – 4 km

1985-present

X

X

X X

The proposed data set in this study will incorporate more careful post-processing QC than

10


any other previously derived global LSWT data set (Table 1). It will utilize the recently improved the Pathfinder AVHRR-derived 4 km resolution LSWT from 1985-2014 with the high resolution MODIS radiometer SST retrievals available at 1 km globally since 2000. By applying rigorous quality control, residual cloud screening, and statistical tests to these data sets, as well as conducting rigorous validation of the resulting data set at dozens of lakes worldwide for long-term periods, the derived LSWT climatology will provide a much more robust data set of LSWT for use in climate change or regional climate simulation studies. Extensive gaps in the availability of clear-sky satellite thermal retrievals due to persistent and highly variable seasonal cloud cover over many of the mid-latitude regions of the earth where lakes are abundant makes it difficult to obtain representative LSWT analyses on a daily basis (Politi et al. 2012; Fiedler et al. 2014). Consequently, it is not surprising that the vast majority of satellite-derived LSWT climatological trend studies have focused on the less-cloudy summer season (e.g., O’Reilly et al. 2015; Torbick et al. 2016).

Figure 3. Mean lake surface temperature for the Great Salt Lake, Utah USA for a 15 day period (1-16 October 2015) period using NOAA Coastwatch AVHRR-3 geostationary satellite data. (a) after careful visual removal of imagery with cirrus cloud contamination b) The same period with no manual removal of suspect data. The Clouds from AVHRR (CLAVRR) cloud masking algorithm is applied in both cases but was ineffective in removing thin cirrus. The GSL is bisected by a railroad causeway which severely restricts water excahnge in the center of the lake, resulting in the large temperature gradient observed.

Over some lakes, available observations during month-long periods may only be one or two satellite images. Thus, for climate-derived trends, great care needs to be taken when using these

11


sparse measurements to determine climate forcing at lakes. In this study, we will carefully analyze the frequency and number of these gaps, and if they have any impacts on climatological trends. To summarize the goals of this study, we will use cloud clearing and statistical quality control algorithms for use with satellite-derived lake temperature datasets from two state-of-the-art global climate SST products (NOAA Pathfinder V5.3 (PFV53) and NASA MODIS LST). The MODIS data set, with 17 years of global retrievals with high spatial resolution (1 km) will be critical for obtaining the first global climatology of LSWT for small lakes (< 20 km in diameter). The NOAA Pathfinder V5.3 has the following improvements that make this version amenable to utilizing for satellite-derived LSWT:

The SSTs in PFV53 are now available for all quality levels, including quality '0'(often lake SST data is lower quality than oceans and lost when 0 is not included)

The PFV53 land mask has been updated (based on Global Lakes and Wetlands Database: Lakes and Wetlands Grid Level 3, 2015) (better boundaries for lakes)

PFV53 also includes L2P (Level 2 Pre-processed) and L3U (Level 3 Uncollated) product levels for the very first time (having this data will assist in the in situ validation)

This study will culminate in the production and archival (within the NOAA National Centers for Environmental Information (NCEI) repository) of new derived data sets LSWT climatological annual cycles in LSWT and trend daily time series with associated error and uncertainty estimates for use by global climate change modeling and observational studies

III. Methods1. Study Design

The proposed research will be led and conducted by PI Crosman at the University of Utah with assistance from a M.S. student 100% dedicated to this project. The proposed project timeline is two years and deliverables will have immediate application to the research and public community. We will rely on the extensive data archives and computational resources at the University of Utah and the 30 tybyte data storage capabilities that will be purchased from the University of Utah Center for High Performance Computing for this study.

The following sections will expand on our study design. First, we will assemble and subset the satellite and in situ LSWT datasets at the outset of the study. Second, an initial validation of the satellite data sets in comparison to in situ data set will be conducted to identify the frequency of cloud contamination and the average length of gaps in the coverage, which could result in sampling errors in monthly or seasonal satellite-derived LSWT trend analyses. Then, informed by this analysis and a literature survey (e.g., Grim et al. 2013) we will develop quality-control algorithms for processing time series of the data at each lake. All available techniques will be utilized to attempt to remove problematic retrievals. The differences in the error statistics of the preliminary and quality-controlled LSWT data sets will be used to quantify the potential impacts of these errors on climate trend analysis. Finally, the highest-quality possible climate data set will be produced for use in forcing numerical weather climate models, as well as a long time series of LWST at each lake. The number of lakes to be included in this analysis in uncertain at this time, and will depend on the quality of the underlying data set, but we expect several hundred small lakes (that have not been previously processed in this manner) and a number of larger lakes as well.

12


2. Assemble global LSWT data sets The first task of this study will be to download the data sets to be used (Table 2). We will download the entire record of the recently reprocessed new and improved global NOAA pathfinder V3.5 SST data set at ~4 km resolution between 1984-2014 at ftp://ftp.nodc.noaa.gov/pub/data.nodc/pathfinder/Version5.3. This NOAA Pathfinder V5.3 data set includes improved lake shorelines as well as associated files that include the individual times of each satellite pass and associated data that went into the daily files. This information was not available to an earlier study that utilized an earlier version of Pathfinder (Schneider and Hook 2010; O’Reilly et al. 2015). In addition, lower-quality data that has historically been thrown out of the Pathfinder retrievals (GHRSST level 0) is now included in the data set. Crosman and Horel (2009) found that a major limitation of the older Pathfinder data set was a lack of data due to only higher quality pixels being included. Since lakes often do not pass ocean-designed QC checks, presumably good LSWT data over lakes was removed from the older version. The level 0 data in the new Pathfinder V3.5 data set will be cautiously analyzed to determine if it can be utilized to increase temporal coverage over some lakes.For small lakes that will be unresolved by the NOAA Pathfinder V3.5, the MODIS LST or SST daily data sets will be downloaded from one of the many sources listed at: https://lpdaac.usgs.gov/dataset_discovery/modis/modis_products_table/mod11a1_v006Or https://modis.gsfc.nasa.gov/data/dataprod/mod28.php . An initial processing methodology described in Crosman and Horel (2009) will be implemented to subset daily files to obtain data from 2000-2016 for small lakes over many regions of the world.

Table 2. SST products to be downloaded and processed in deriving quality-controlled LSWT data set, as well as ancillary data sets to be utilized

Data Set Resolution(if applicable)

Period of Record

Source

NOAA Pathfinder V5.3 ~4 km 1981-2014 NOAAMODIS SST/LST ~1 km 2000-present NASAIn situ buoy and sampling data for validation and cloud contamination and gap error analysis

1985-present 1.Mesowest.utah.edu2.Sources listed in Sharma et al. (2015) 3. Additional lake-specific sources

Inland water dataset for distance-to-land, distance-to-water, water-body identifier and lake-centre co-ordinates

Circa 2012 Carrea et al. 2015

Numerous in situ data sets have already been obtained by the PI for a number of lakes in North America. We will utilize the in situ lake temperature data from global lakes available for download in an excel file by Sharma et al. 2015. We will also utilize contact information provided in the global data set of Sharma et al. (2015) to contact for updated and more temporally specific in situ data sets from global lakes to develop a large and long time series of

13

https://modis.gsfc.nasa.gov/data/dataprod/mod28.php

https://lpdaac.usgs.gov/dataset_discovery/modis/modis_products_table/mod11a1_v006

ftp://ftp.nodc.noaa.gov/pub/data.nodc/pathfinder/Version5.3


in situ data sets to compare against the satellite observations. In addition, we will download all available National Buoy Data Center (NDBC) retrievals and any other buoy data sets that are routinely injested into Mesowest (mesowest.utah.edu) (Horel et al. 2002). An inland water dataset for distance-to-land, distance-to-water, water-body identifier and lake-centre co-ordinates will also be utilized in determining lake-specific quality control algorithms.

3. Validation, Analysis and Lake Climatology Data Product Development Plana. In situ versus LSWT validation at dozens of lakes worldwideWe will initially determine temporal coverage of available satellite observations for each quality level of the satellite data sets. Second, we will conduct a validation of all available lakes where in situ data has been obtained to determine the quality of the Pathfinder V5.3 and MODIS LST data sets over long time periods as a function of year, month, and day. A key shortcoming of previous global processing of satellite-derived LSWT has been limited in situ validation confined to several (mostly) large lakes, such as the Laurentian Great Lakes in North America. In addition, long-term validation has been less common, but of interest for climatological data sets. Most satellite-derived LSWT studies compared the satellite data against in situ data using short-term case study validation periods covering months to at most several years of satellite vs in situ measurements. To our knowledge, no rigorous, inter-annual validation over multiple lakes worldwide has ever been conducted previously (Crosman et al. 2017b). For example, Crosman et al. (2017a) found that seasonal and interrannual variations in in situ versus satellite-derived LSWT retrievals were noted for several lakes in the USA, potentially due to greater cloud contamination and gap errors in the springtime versus summer and fall in the 5-day blended analysis used in that study (Fig. 4).

b. Evaluation of impacts of residual cloud contamination on satellite-derived LSWT and develop lake-specific QC procedures

A key component of the QC procedures will involve first utilizing several decades of climatological average LSWT for each lake to help inform realistic ranges in annual, seasonal, and daily LSWT variability. The in situ versus satellite-derived LSWT validation statistics for the NOAA Pathfinder V3.5 and MODIS LST

14

Figure 4. From Crosman et al. (2017). In situ buoy data versus satellite-derived LSWT analysis for Lake Michigan USA from 2007-2014.


data sets will then establish a preliminary spatial and temporal LSWT climatology for use in QC algorithm climatological checks to be applied based on the large multi-year climatology of satellite LSWT data to derive spatial plots of mean LSWT over lakes that are large enough. For the small lakes where insufficient pixels exist for determination of spatial gradients in LSWT, the climatological time series of LSWT will inform the typical daily changes in LSWT observed. Using this preliminary satellite-derived climatology, unrealistic variations in LSWT, either spatially or temporally, will help flag cloud contaminated images that miss detection through standard cloud masking algorithms. Some of the QC algorithms likely to be utilized in this study are listed in Table 3. We will evaluate both LSWT and cloud detection satellite products for several months of data for each lake to determine the best spatial heterogeneity or other statistical tests to apply to remove or flag the periods with residual cloud contamination.

Careful analysis of spatial and temporal characteristics of satellite-derived LSWT images will be used to assist in developing simple spatial heterogeneity and other techniques to

Table 3. Preliminary proposed LSWT quality-control tests

Name DescriptionLake-specific spatial heterogeneity test

Determine typical allowable spatial gradients based on bathymetry and climatology

Lake-specific daily consistency test

Flag data as suspect if daily change in LSWT is larger than determined threshold for a given lake

Lake-specific climatological tests

Flag data as suspect if LSWT falls outside of climatological range

Other tests To be developed per available techniques in the literature and upon analysis of the shortcomings between in situ and satellite-derived data sets. These may include tests to screen for dust or smoke from persistent wildfires

identify unphysical gradients in LSWT resulting from cloud contaminated pixels. The cloud screening techniques will initially use both visual inspection and climatological consistency checks, following statistical approaches similar to Grim et al. (2013) to remove residual cloud contaminated imagery that was not removed by the Pathfinder or MODIS cloud screening algorithms. We will also conduct a thorough literature survey on any other QC techniques that may be effective and potentially utilize more sophisticated cloud screening products as well. After cloud contamination screening methods have been evaluated using in situ observations and visual inspection on an initial subset of lakes, the algorithms will be applied without time-consuming visual inspection of the imagery to a greater number of lakes. As time and resources allow, case studies of the impacts of smoke from persistent wildfires, or of widespread dust storms on LSWT retrievals will be conducted, with the goal of quantifying the potential impact of these sources on monthly to seasonal averages of LSWT. Ice cover is frequently observed in many high latitude lakes during the wintertime. Ancillary data sets (such as daytime visible MODIS reflectance) will be identified to determine and flag the ice covered periods on each lake in this study. Non-saline lake temperatures below 0 ᵒC will automatically be flagged as ice.

c. Determine impacts of residual cloud contamination on satellite-derived LSWT trends

15


After adequate cloud screening procedures have been determined, we will compare the reduction in the number of available images at each lake due to the QC algorithms, and the differences in the before and after QC and cloud-cleared satellite LSWT imagery calculated trends in the LSWT time series. While a number of studies have introduced the concept of these errors impacting LSWT trend studies, to our knowledge no previous study has attempted to quantify the impact of cloud contamination on satellite-derived LSWT trends.

d. Quantify impacts of temporal gaps in coverage on climatological LSWT trendsFollowing the application of the QC algorithms, we will conduct a careful analysis of the frequency of available cloud-free images over each lake. The impact of ‘gaps’ in lake coverage resulting from common cloudy periods (particularly in cool seasons in mid-latitudes) on the ability to determine climate trends at that lake will be a function of lake depth, latitude, and other geophysical forcing mechanisms that control the variability of LSWT for a given lake. To determine the impact of gaps in coverage on climatological satellite-derived LSWT trends, we will rely heavily on comparisons with in situ data from a representative sampling of lakes worldwide described earlier. The frequency of satellite images needed to adequately describe LSWT for climate studies is expected to vary between lakes. For example, over a deep lake in the tropics, such as the African rift lakes, the changes in LSWT over a few week period is likely much smaller than over, for instance a shallow continental lake such as Utah’s Great Salt Lake, where lake temperature over a cloudy period may decrease by several °C per day (Crosman and Horel 2009). Thus, a single image over a 3-week period may be sufficient to represent the temperature within acceptable uncertainty in the African rift lakes, but introduce a sampling error larger than 5 °C on a lake such as the Great Salt Lake. Thus, a key finding of this task will be to determine

16

Figure 5. Annual time series of daily lake temperature (ᵒC) for 2007-2015 for (a) Lake Michigan, (b) Lake Okeechobee, (c) Lake Oneida derived from NASA MUR daily analyses from Crosman et al. (2017a).


where cloud cover is so high that the infrequent available images are insufficient to characterize the LSWT and to remove those periods and lakes from the final available LSWT climatology.

e. Development and archival of new long-term daily time series and climatology of LSWT for climate trend analysis

We propose to develop a climatology similar to Fig. 5 (black line) for up to 1000 lakes worldwide (the exact number will depend on the outcome of the cloud gap and data availability analysis). The only climatology currently available is the ARC-Lake dataset which utilizes less-frequently available AATSR satellite data. In the new dataset we will develop, climatological daily (when available) satellite-derived LSWT records extended to many smaller lakes worldwide. The previously mentioned datasets (Pathfinder V3.5 and MODIS LST) combined with the cloud screening, statistical, and other quality control procedures will result in the best available new long-term daily time series and climatology of LSWT for climate trend analysis that is a critical need for global regional climate change analysis and for input into high-resolution regional climate simulations. Two data sets will result from processing the quality-controlled Pathfinder V3.5 and MODIS LST global satellite-derived LSWT data:

1). Time series of daily satellite-derived LSWT for the period of record (1985-2014 for Pathfinder V3.5, 2000-2017 for MODIS LST) will be created for each lake over each time period that was found to have sufficient retrievals. For lakes between 4 and 16 km in diameter, the quality-controlled MODIS LST will be the primary data source. For lakes greater than 16 km in diameter, both MODIS LST and Pathfinder V3.5 LSWT will be utilized. 2). Weekly annual climatology of satellite-derived LSWT. This derived data set will be the mean weekly temperature of each lake over the respective satellite periods of record (both MODIS LST and Pathfinder V3.5 LSWT records will be computed separately on lakes large enough to be resolved by both data sets).

IV. Project Duration and Timeline

Project Milestone Description Delivery Dates

Collect climatological LSWT data

Assemble global satellite-derived LSWT data sets (e.g., NOAA Pathfinder V5.3, MODIS LST).

Summer 2018

In situ versus LSWT validation at dozens of lakes worldwide

Perform rigorous in situ validation of satellite retrievals for numerous lakes over long periods of record, which has not been previously conducted.

July 2019

Create preliminary satellite-derived LSWT climatology

The preliminary LSWT climatology will be produced before and QC and improved cloud screening procedures. This climatology will provide constraints to be used in QC algorithms to improve the final LSWT climatology

December 2019

Apply cloud masking and statistical QC

Develop and apply cloud masking and statistical QC algorithms to remove bad data from LSWT

July 2019

17


algorithms data sets.Provide recommendations on potential impact of LSWT errors on climate trend analysis

Quantify the potential impact of errors and uncertainties associated with both cloud contamination and temporal gaps (long cloudy periods) on climatological satellite lake temperature trends

June 2020

Provide archived publically-available new long-term daily time series of LSWT for climate trend analysis

Upload new quality-controlled LSWT climatological time series data sets for global (including small to medium-sized lakes not incorporated in previous climatologies) with associated error and uncertainty estimates to a publically-accessible archive

June 2020

Provide archived publically-available new long-term climatology of LSWT for climate studies and initializing climate models

Upload new climatological LSWT data sets (weekly mean climate values over annual diurnal cycle) for global (including small to medium-sized lakes not incorporated in previous climatologies)

June 2020

V. Project Compliance - NEPA and Other ClearancesNo NEPA or other clearances are required for this project.

VI. Research LinkageParticipation of the MesoWest team in the National Mesonet Program supports the acquisition, archival, and dissemination of observations, including in situ buoy lake temperature observations on lakes from many locations in the CONUS. The goals of this study will allow preliminary research conducted during the NASA MISST project which evaluated error sources of LSWT and evaluated real-time satellite-derived analyses LSWT for numerical weather prediction to be extended in this study to a larger project evaluating climate-length LSWT retrievals and needs. The PI is also involved in a NOAA JFSP numerical modeling study over the Western US where modeling the flows and humidity around fires depend on accurate specification of lake surface temperature.

Grant Program

Project or Proposal Description/Identification

Funding Amount

PI Completion Date

NOAA/NWS National Mesonet Project 750,000 Co-PI Crosman

Sept. 2019

NASA Multi-sensor Improved Sea Surface Temperatures

120,00 Crosman Dec. 2017

JFSP Assessment of HRRR Model Forecasts of Convective Outflows in

the Fire Environment

277,068 Co-PI Crosman

June 2019

VII. Deliverables Project deliverables include: (1) results from all phases of the analysis of errors associated with cloud contamination and gaps in LSWT and (2) new publically-accessible quality-controlled

18


time series and climatological LSWT data for global lakes. We will communicate updates to NASA, NOAA, NWS and academic colleagues who are part of the Group for High Resolution Seas Surface Temperature (GHRSST) community and will be called upon to advise this research and receive updates on project progress from the research team. Project results will be described and summarized in both the JFSP report and 2 submitted journal publications. A M.S. thesis will result from this project. Conference presentations will be given at the GHRRST annual meeting and American Geophysical Union (AGU) annual meeting and AGU Ocean Science meeting which is held in concert with the American Society for Limnology and Oceanography (ASLO). As part of this study, rigorous public outreach activities will also be conducted. The PI will work with the VP for research office at the University of Utah to communicated research findings via social media, and the PI will also prepare Presentations to local school groups and other educational entities.

Deliverable Type Description Delivery Dates

Archive of improved climate LSWT data sets

Develop repository of quality-controlled satellite-derived climatological LSWT time series and mean climatological data sets

June 2020

Conference presentation

Present research results at GHRSST and AGU, AGU Ocean Sciences, and American Limnological Society

Journal articles Submit findings on impact of errors and uncertainties associated with cloud contamination and temporal gaps (long cloudy periods) on climatological satellite lake temperature trends. Also submit another journal article to nature scientific data journal outlining the LSWT climate data sets

May 2020

MS thesis Evaluation of climate trends in global small lakes from satellite

June 2020

Final report Submission of final report including recommendations to improve communication of outflow forecasts to fire personnel

July 2020

VIII. Roles of Investigators and Associated Personnel Personnel Role Agency ResponsibilityErik Crosman PI University of Utah Overall project responsibility, lead

all and conduct all scientific work in concert with M.S. student

19


Data Sharing PlanI. Data Management Plan JustificationThe proposed project will result in the acquisition and temporary storage of between 20 and 30 tbytes of global satellite SST files in netcdf and hdf format, downloaded from the daily global NOAA pathfinder V3.5 SST data set between 1984-2014 at ftp://ftp.nodc.noaa.gov/pub/data.nodc/pathfinder/Version5.3 and the MODIS LST or SST daily data sets downloaded from one of the many sources listed at: https://lpdaac.usgs.gov/dataset_discovery/modis/modis_products_table/mod11a1_v006Or https://modis.gsfc.nasa.gov/data/dataprod/mod28.phpThe derived climatological data sets and associated time series for each of the several hundred lakes that will be processed as part of this study will be permanently archived. The research group that Dr. Crosman is part of at the University of Utah has extensive experience managing large data archives obtained from real-time monitoring sites as well as field programs. For example, MesoWest software (http://mesowest.utah.edu) accesses, archives, and makes available environmental observations from over 40,000 surface stations (http://mesowest.org/api). We have also recently started a publically-accessible archive for the High Resolution Rapid Refresh Model http://home.chpc.utah.edu/~u0553130/Brian_Blaylock/cgi-bin/hrrr_download.cgi

II. Project Data Management1. Environmental Data types-NOAA pathfinder V3.5 SST: These data files (in netcdf format) between 1984-2014 will be downloaded for each day. These daily files (daytime and nighttime are both produced daily) are each of ~37 Mb is size. Thus, calculating ~37 Mb per day * 365 days per year * ~30 years (1985-2014) results in ~1 tbyte of data. -MODIS LST or SST: Depending on how many global swaths of higher-resolution MODIS imagery is downloaded, these files will result in up to 4.8 Gb of data per day * 365 days per year * ~17 years results in ~27 tbytes. Because we do not need to extract all global regions, this is an upper estimate and the expected size would likely be below this, and if needed we will sub-set lake regions and not store all of the large files. -In Situ Verification Data Sets: These data sets are small and generally consist of csv, text, or excel spreadsheet files and are not expected to result in any notable storage requirements. -Ancillary data sets: These will include data sets to help aid in the quality control of the LSWT data sets, including the Inland water dataset for distance-to-land, distance-to-water, water-body identifier and lake-centre co-ordinates (Carrea et al. 2015). The total storage requirements of these data sets is expected to be less than 1 tbyte.

2. Quality AssuranceNOAA pathfinder V3.5 SST and MODIS SST/LST: These data are widely used and provided in an output format (netcdf and hdf) that insures consistency in the fields. Verification Data Sets: All in situ observations obtained through the MesoWest or Sharma et al. (2015) archive are processed and will be subjected to quality control procedures upon receipt.

3. Data AccessAlmost all of the information we will use in our analysis is in the public domain. Only in situ

20

http://home.chpc.utah.edu/~u0553130/Brian_Blaylock/cgi-bin/hrrr_download.cgi

http://mesowest.org/api

http://mesowest.utah.edu/

https://modis.gsfc.nasa.gov/data/dataprod/mod28.php

https://lpdaac.usgs.gov/dataset_discovery/modis/modis_products_table/mod11a1_v006

ftp://ftp.nodc.noaa.gov/pub/data.nodc/pathfinder/Version5.3


lake validation data sets in original format obtained from individual researchers and lake environmental managers would require permission of the agencies who supplied that information. All of the research to be completed will be undertaken within the secure computing environments of the research teams at the University of Utah that have appropriate security measures in place to limit direct access to the project team. As the project progresses, processed information will be accessible via a web server at the University of Utah for anyone interested in the research to help in the evaluation of the project results.

4. Storage and BackupProcedures are already in place to manage the large data sets received and archived at the University of Utah. The PIs research group has over 100 tbytes of disk storage available in the CHPC computing system, with approximately 15 tbytes currently available. As part of this proposal an additional 30 tbytes of disk storage for this project will be purchased by CHPC, allowing for a total storage for this project of up to 45 tbytes. CHPC provides a robotic tape system that is capable of hosting critical data required for our study and the Principal Investigator has sufficient numbers of tapes to store a backup of all data resources during the project.

III. Long-Term Data Management1. MetadataOur metadata documents will be included with the NOAA National Centers for Environmental Information (NCEI) data repository described below prior to completion of the project conforming to appropriate standards specified by NOAA. Any publications that result from this research and papers describing the data sets will be submitted to the NOAA Institutional Repository after acceptance. We will update earlier metadata provided to NOAA as needed: repository name, URL, and repository-assigned identifier.

2. Data RepositoryThe data repository for the derived climatological lake temperature fields will be the NOAA NCEI. We submitted a contingent request for archiving data in the NCEI in July 2017 and were approved prior to submitting this proposal. We have experience undertaking research that culminates in the underlying data being archived in a data repository (doi:http://dx.doi.org/10.5065/D6028PRS) and having that data described in a peer-reviewed data journal (e.g., Jacques et al. 2016). The Marriott Library at the University of Utah is in the process of developing a data repository designed to meet the needs of researchers who intend to archive data relied upon to complete sponsored research. We have been in contact with the Library’s data repository team to assess whether the resources will be available at the completion of our project to archive the downloaded raw Pathfinder and MODIS data evaluated in this study (this data will not be archived at NOAA NCEI, only the quality controlled derived climatology). Because of the potential high interest in the data for further research, we will pursue describing the resulting data set in a publication in a data science journal, such as that by Sharma et al. (2015).

3. Data AccessAll of the data permanently archived at NOAA NCEI or the Marriott Library data repository will be classified as “open access” and will be publically available free of charge within the two year

21


time window required by NOAA.

Budget Justification

PERSONNEL

PI – Erik Crosman (effort = 6 cal. mos.): Dr. Crosman will direct the overall project operation and as a research assistant professor will be directly involved in all technical and data processing, quality control, and archiving aspects of the project, including downloading, processing, and developing the satellite-derived lake temperature climatology and temporal time series.

Graduate Student #1 (effort = 12 cal. mos): This position will involve a M.S. student who will commit 100% effort to this project, subsetting, processing, and developing the satellite lake temperature data sets, and applying the various quality control and other processing techniques related to this study. Anticipated hire date 20 August 2018.

Salary increase of 4% is included in Year 2.

FRINGE BENEFITS

The fringe benefit rate for full-time faculty/staff is calculated at 37% for all years. The fringe benefit rate for graduate students is calculated at 8% for all years. Salary increase of 4% is included in Year 2.

TRAVEL – DOMESTIC Total: $10,000Travel funds for Year 1 include support for both the PI and graduate student to participate in one annual remote sensing or global change conference per year. Funds are requested in Year 1 for the graduate student to attend the ASLO Aquatic Sciences Meeting and for the PI to attend the Group for High Resolution Sea Surface Temperature (GHRSST) meeting in Year 1.

Year 1:ASLO Aquatic Sciences Meeting:

1 trip x 1 persion @ $750 airfare: $750Ground transportation: $1656 nights hotel @ $140/night: $8405 days per diem @ $69/day: $345Conference registration $500Total $2600

GHRSST annual meeting: 1 trip x 1 persion @ $750 airfare: $850Ground transportation: $1656 nights hotel @ $140/night: $8405 days per diem @ $69/day: $345Conference registration $200Total $2400

Travel funds for Year 2 include support for both the PI and graduate student to participate in one annual remote sensing or global change conference determined to be the most appropriate and effective way to disseminate the study results to the scientific community. The PI will present the results of this study at the American Geophysical Union Ocean Sciences (AGU OS) joint meeting with the Association for the Sciences of Limnoloigy and Oceannography (ASLO) (location to be determined) and results will also be presented by the graduate student at the AGU Fall Conference in San Francisco, CA.

Year 2:AGU OS/ASLO meeting:

1 trip x 1 person @ $550 airfare: $5505 days rental car: $300

22


6 nights hotel at $160/night: $9605 days per diem at $69/day: $345Conference registration: $500Total: $2655

AGU Fall Meeting:1 trip x 1 person @ $350 airfare: $350Ground transportation: $1505 nights hotel at $200/night: $10005 days per diem at $69/day: $345Conference registration: $500Total: $2345

PUBLICATIONSSupport is requested for one publication in Year 2 ($3000 per manuscript).

SUPPLIESMaterials and supplies (peripheral computer hardware, backup tapes, etc.): Year 1 $499, Year 2 $493.

OTHER30 tbytes of disk storage for processing large global lake temperature data sets (100% used by project): $5000 in Year 1.

INDRECT COSTSUniversity of Utah indirect costs are calculated at a rate of 52.5% of a Modified Total Direct Cost (MTDC).

23


Erik T. CrosmanResearch Assistant Professor, Department of Atmospheric Science, University of Utah

[email protected], 505 570-0552 135 S 1460 East Rm 819 Salt Lake City, UT 84112-0110

EDUCATION* Ph.D. 2011, Atmospheric Sciences, University of Utah* M.S. 2005, Atmospheric Sciences, University of Utah* B.S. 2003, Earth Science in Meteorology, University of Northern Colorado

PROFESSIONAL EXPERIENCE*2015-present, Assistant Research Professor, University of Utah *2011-2015, Postdoctoral Research Assistant, University of Utah *2003-2011, Research Assistant, University of Utah RELATED RESEARCH ACTIVITIESMy research interests fall broadly in four key areas: 1) Remote sensing of lakes, 2) thermally-driven flows with an emphasis on lake and sea-induced circulations, 3) numerical modeling, and 4) meteorological and air quality observations in complex mountainous terrain. I am currently working on evaluating daily lake temperature blended analyses from the NASA MUR product for the purpose of utilizing the data for input into regional and climate models. I am also working with the Utah Division of Air Quality to improve the meteorological simulations of winter stable layers for input into air quality models. Other active projects include understanding the impact of thermally-driven flows in the Salt Lake Valley on ozone and PM2.5 concentrations and transport, developing an observational platform to measure ozone and particulate sensors on light rail trains in an urban environment.

PUBLICATIONS IN LAST 3 YEARSCrosman, E., and coauthors, 2017: Satellite-derived lake surface water temperature: A review.

Submitted, Remote Sensing of Environment.Crosman, E., J. Vazquez-Cuervo, M. Chin, 2017: Evaluation of the Multi-Scale Ultra-High

Resolution (MUR) analysis of lake surface temperature. Remote Sensing. 9(7), 723. doi:10.3390/rs9070723

Crosman, E., J. Horel, 2017: Large-eddy simulations of a Salt Lake Valley Cold-air Pool. Atmospheric Research. 193, 10-25.

Crosman, E. T., Jacques, A. A., & Horel, J. D., 2017: A novel approach for monitoring vertical profiles of boundary-layer pollutants: Utilizing routine news helicopter flights. Atmospheric Pollution Research. 8, 828-835.

Foster, C., E. Crosman, J. Horel, 2017: Simulations of a Cold-Air Pool in Utah’s Salt Lake Valley: Sensitivity to Land Use and Snow Cover. Boundary Layer Meteorology. 164, 63-87.

Blaylock, B., J. Horel, E. Crosman, 2016: Impact of Lake Breezes on Summer Ozone Concentrations in the Salt Lake Valley. J. Appl. Meteor. Clim. 56, 353-370.

Horel, J., E. Crosman, A. Jacques, B. Blaylock, S. Arens, A. Long, J. Sohl, R. Martin, 2016: Influence of the Great Salt Lake on summer air quality over nearby urban areas. Atmospheric

24


Science Letters. 17, 480-486. Jacques, A., J. Horel, E. Crosman, F. Vernon, J. Tytell, 2016: The Earthscope US Transportable

Array 1 Hz Surface Pressure Dataset. Geoscience Data Journal, 3: 29–36. Crosman, E., and J. Horel 2016: Winter lake breezes near the Great Salt Lake. Boundary Layer

Meteorology. 159, 439–464. doi:10.1007/s10546-015-0117-6Jacques, A., J. Horel, E. Crosman, F. Vernon, 2015: Central and Eastern United States surface

pressure variations derived from the USArray network. Mon. Wea. Rev. 143, 1472-1493. Lehner, M., C. David Whiteman, S.W. Hoch, E.T. Crosman, M.E. Jeglum, N.W. Cherukuru,

R. Calhoun, B. Adler, N. Kalthoff, R. Rotunno, 2015: The METCRAX II field experiment—A study of downslope windstorm-type flows in Arizona’s Meteor Crater. Bull. Amer. Meteor. Soc., 97, 217-235.

Neemann, E., E. Crosman, J. Horel, L. Avey, 2015: Simulations of a cold-air pool associated with elevated wintertime ozone in the Uintah Basin, Utah. Atmos. Chem. Phys., 15, 135-151.

Strong, C., A.K. Kochanski, E. Crosman, 2015: A Slab Model of the Great Salt Lake for Regional Climate Simulations. J. Adv. Model Earth Systems, 6, 602-615.

5 OTHER SELECTED PUBLICATIONSLareau, N., Crosman, E., Whiteman, C.D., Horel, J.D., Hoch, S.W., Brown, W.O.J., Horst, ,

2013: The Persistent Cold Air Pool Study, Bull. Amer. Meteor. Soc., 94(1), 51-64Grim, J.A., J.C. Knievel, E. Crosman, 2013: Techniques for Using MODIS Data to Remotely

Sense Water Surface Temperatures. Journal of Atmospheric and Oceanic Technology, 30, 2434-2451.

Crosman, E.T., J.D. Horel, 2012: Idealized Large-Eddy Simulations of Sea and Lake Breeze: Sensitivity to Lake Diameter, Heat Flux, and Stability. Boundary Layer Meteorology. 144, 309-328.

Crosman, E.T., J.D. Horel, 2010: Sea and lake breezes: A review of numerical studies. Boundary Layer Meteorology. 137, 1-29.

Crosman, E.T., J.D. Horel, 2009. MODIS-derived Surface Temperature of the Great Salt Lake. Remote Sensing of Environment. 113, 73-81.

SELECTED EDUCATIONAL AND PROFESSIONAL ACTIVITIES IN PAST THREE YEARS* Supervising 1 Ph.D. student presently* On Supervisory committees of 4 student M.S. thesis and 1 student Ph.D. thesis* Co-Chair, 17th AMS Mountain Meteorology Conference, Burlington VT. 2016.*Faculty advisor, U. of Utah Student Chapter of the American Meteorological Society. 2016- *Instructor, ATMOS 5050/6050: Atmospheric Instrumentation. 2014, 2016. *Team Member, Cold Pool Modeling Working Group, Environmental Protection Agency (EPA). 2015- *Team member, NASA Multi-sensor Improved Sea Surface Temperatures (MISST). 2011-AWARDS* Edward Zipser Outstanding Graduate Student Award, University of Utah. 2011* Outstanding Service and Leadership Award for the Persistent Cold Air Pool Study (PCAPS). 2011* Best Student Oral Presentation award at the 14th Conference on Mountain Meteorology. 2010*NASA Earth Systems Science Fellowship Recipient. 2006-2009 *2006-2009 College of Mines Outstanding Teaching Assistant Award, 2006

25


1. Current and Pending Support

Erik T. Crosman

Current Support1) Title: Multi-sensor Improved Sea Surface Temperatures (MISST2): 2012-2017Supporting Agency: NASATotal Award Period: 11/15/2012-11/14/2017Total Award Amount: 127,000Annual Man Months: 1.52) Title: UUNET DATA PURCHASE FOR NWS NATIONAL MESONET PROGRAMSupporting Agency: Synoptic Data CorporationTotal Award Period: 01/17/2017-01/16/2018Total Award Amount: 33,600Annual Man Months: 0.52) Title: Assessment of HRRR Model Forecasts of Convective Outflows in the Fire EnvironmentSupporting Agency: Department of Interior (Bureau of Land Management)Total Award Period Covered: 08/01/2017-07/31/2019Total Award Amount: $277,068Annual Man Months: 3

Pending Support:1) Title: Improving Long-term Records of High-Resolution Satellite-derived Lake Surface Temperature for Global and Regional Climate StudiesSupporting Agency: NOAATotal Award Period Covered: 7/01/2018-6/30/2020Total Award Amount: $268,424Annual Man Months: 6

26


References

Adrian, R.A., O’Reilly, C.M., Zagarese, H., Baines, S.B., Hessen, D.O., Wendel, K., Livingstone, D.M., Sommaruga. R., Dietmar, S., Van Donk, E., Weyhenmeyer, G.A., & Winderl, M. (2009). Lakes as sentinels of climate change. Limnol. Oceanogr.,54(6, part 2), 2283–2297.Balsamo, G., Salgado, R., Dutra, E., Boussetta, S., Stockdale, T. and co-authors. (2012). On the contribution of lakes in predicting near-surface temperature in a global weather forecasting model. Tellus A. 64, 15829. Cael, B. B., A. J. Heathcote, & Seekell, D.A. (2017). The volume and mean depth of Earth's lakes, Geophys. Res. Lett., 44, 209–218.Castendyk, D.N., Obryk, M.K., Leidman, S.Z., Gooseff, M., Hawes, I. (2016). Lake Vanda: A sentinel for climate change in the McMurdo Sound Region of Antarctica. Global and Planetary Change, 144, 213–227.Crosman, E. T., & Horel, J.D. (2009), MODIS-derived surface temperature of the Great Salt Lake, Remote Sensing of Environment, 113, 73–81.Crosman, E., J. Vazquez-Cuervo, M. Chin, 2017a: Evaluation of the Multi-Scale Ultra-High Resolution (MUR) analysis of lake surface temperature. Remote Sensing. 9(7), 723. doi:10.3390/rs9070723Crosman, E., and coauthors, 2017b: Satellite-derived lake surface water temperature: A review. Submitted, Remote Sensing of Environment.Dörnhöfer, K., Oppelt, N. (2016). Remote sensing for lake research and monitoring – Recent advances. Ecological Indicators, 64, 105-122. Dutra, E., Stepanenko, V. M., Balsamo, G., Viterbo, P., Miranda, P. M. and co-authors. 2010. An offline study of the impact of lakes on the performance of the ECMWF surface scheme. Boreal Environ. Res. 15, 100–112.Fan, X., Tang, B.-H., Wu, H., Yan, G., Li, Z.-L. (2015), A three-channel algorithm for retrieving night-time land surface temperature from MODIS data under thin cirrus clouds. International Journal of Remote Sensing, 36, 19-20. Fiedler, E., Martin, M. & Roberts-Jones, J. (2014). An operational analysis of lake surface water temperature. Tellus A. 66, 21247. Foster, C., E. Crosman, J. Horel (2017). Simulations of a Cold-Air Pool in Utah’s Salt Lake Valley: Sensitivity to Land Use and Snow Cover. Boundary Layer Meteorology. 164, 63-87.Grim, J.A., Knievel, J.C., & E.T. Crosman (2013). Techniques for using MODIS data to remotely sense lake water surface temperatures. Journal of Atmospheric and Oceanic Technology, 30(10), 2434-2451.Hook, S.J., , R. C., MacCallum, S. and Merchant, C. J. (2012). Lake surface temperature [in “State Absolute of the Climate in 2011”] Bulletin of the American Meteorological Society, 93 (7). S18-S19. Horel, J., Splitt, M., Dunn, L., Pechmann, J., White, B., Ciliberti, C., ... & Burks, J. (2002). Mesowest: Cooperative mesonets in the western United States. Bulletin of the

27

http://centaur.reading.ac.uk/view/creators/90005270.html


American Meteorological Society, 83(2), 211-225.Hulley, G. C. (2009). MODIS cloud detection over large inland water bodies: Algorithm theoretical basis document. Pasadena, California: Jet Propulsion Laboratory, California Institute of Technology.Hulley, G.C., Hook, S.J., & Schneider, P. (2011). Optimized split-window coefficients for deriving surface temperatures from inland water bodies, Remote Sensing of Environment, 115, 3758-3769.Javaheri, A. Babbar-Sebens, M., & Miller, R.N. (2016). From skin to bulk: An adjustment technique for assimilation of satellite-derived temperature observations in numerical models of small inland water bodies. Advances in Water Resources, 92, 284-298. Kheyrollah Pour, H., Rontu, L., Duguay, C. R., Eerola, K. & Kourzeneva, E. (2014). Impact of satellite-based lake surface observations on the initial state of HIRLAM. Part II: analysis of lake surface temperature and ice cover. Tellus A. 66, 21395.Kraemer, B. M., Chandra, S., Dell, A. I., Dix, M., Kuusisto, E., Livingstone, D. M., Schladow, S. G., Silow, E., Sitoki, L. M., Tamatamah, R. & McIntyre, P. B. (2016). Global patterns in lake ecosystem responses to warming based on the temperature dependence of metabolism. In press, Glob Change Biol. Layden, A., Merchant, C. and MacCallum, S. (2015). Global climatology of surface water temperatures of large lakes by remote sensing. International Journal of Climatology, 35 (15). pp. 4464-4479. Li, Z.L., Tang, B.H., Wu, H., Ren, H., Yan, G., Wan, Z., Trigo, I.F., Sobrino, J.A. (2013) Satellite-derived land surface temperature: current status and perspectives. Remote Sens. Environ., 13, 14–37.MacCallum, S. N., and Merchant, C. J. 2011, ARC-Lake Algorithm Theoretical Basis Document – ARC-Lake v1.1, 1995-2009 [Dataset], The University of Edinburgh, School of GeoSciences/European Space Agency, http://hdl.handle.net/10283/88.MacCallum, S.N., & Merchant, C.J. (2012). Surface Water Temperature Observations of large lakes by optimal estimation. Can. J. Remote Sensing, 38(1), 25 – 45.MacKay M. D. , Neale P. J. , Arp C. D. , De Senerpont Domis L. N. , Fang X. , Gal G. , Jöhnk K. D. , Kirillin G. , Lenters J. D. , Litchman E. , MacIntyre S. , Marsh P. , Melack J. , Mooij W. M. , Peeters F. , Quesada A. , Schladow S. G. , Schmid M. , Spence C. , Stokesr S. L. , (2009), Modeling lakes and reservoirs in the climate system, Limnology and Oceanography, 54, 2315.Mason, L.A., Riseng, C.M., Gronewold, A.D. et al. (2016). Fine-scale spatial variation in ice cover and surface temperature trends across the surface of the Laurentian Great Lakes. Climatic Change, 138, 71-83.Merchant C. J., Harris, A. R., Maturi, E. and MacCallum, S. 2005. Probabilistic physically-based cloud screening of satellite infra-red imagery for operational sea surface temperature retrieval, Quarterly. Journal of the Royal Meteorological Society, 131, 2735-2755.Oesch, D,C., Jaquet, J.M., Klaus, R, Schenker, P (2008). Multi-scale thermal pattern monitoring of a large lake (Lake Geneva) using a multi-sensor approach. Int. Journal of Remote Sensing. 29(20), 5785-5808.O’Reilly, C. M., et al. (2015). Rapid and highly variable warming of lake surface

28


waters around the globe, Geophys. Res. Lett., 42, 10,773–10,781.Politi, E, Cutler, M.J., & Rowan, J.S. (2012). Using the NOAA Advanced Very High Resolution Radiometer to Characterize temporal and spatial trends in water temperature of large European lakes. Remote Sensing Environment. 126, 1-11.Reavie, E. D., Sgro, G. V., Estepp, L. R., Bramburger, A. J., Shaw Chraïbi, V. L., Pillsbury, R. W., Cai, M., Stow, C. A. & Dove, A. (2016), Climate warming and changes in Cyclotella sensu lato in the Laurentian Great Lakes. Limnol. Oceanogr.Riffler M., Lieberherr G. & Wunderle S. (2015) Lake surface water temperatures of European Alpine lakes (1989–2013) based on the Advanced Very High Resolution Radiometer (AVHRR) 1 km data set. Earth System Science Data 7, 1–17.Schneider, P., & Hook, S. J. (2010). Space observations of inland water bodies show rapid surface warming since 1985. Geophysical Research Letters, 37.Sharma, S., et al. (2015), A global database of lake surface temperatures collected by in situ and satellite methods from 1985–2009, Sci. Data, 2.Stendera, S., Adrian, R., Bonada, N., Ca˜nedo-Argüelles, M., Hugueny, B., Januschke, K.,Pletterbauer, F., Hering, D., 2012. Drivers and stressors of freshwater biodiversitypatterns across different ecosystems and scales: a review. Hydrobiologia 696,1– 28, http://dx.doi.org/10.1007/s10750-012-1183-0.Torbick, N., Ziniti, B., Wu, S., & Linder, E. (2016). Spatiotemporal lake skin summer temperature trends in the Northeastern United States. Earth Interact., 20, 1–21.Verpoorter, C., Kutser, T., Seekell, D.A., & Tranvik, L.J. (2014). A global inventory of lakes based on high-resolution satellite imagery, Geophys. Res. Lett., 41, 6396– 6402.Wan, W., Li, H., Xie, H., Hong, Y., Long, D., Zhao, L., Han, Z., Cui, Z., Cui, Y., Liu, B., Wang, C., Yang, W. (2017). A comprehensive data set of lake surface water temperature over the Tibetan Plateau derived from MODIS LST products 2001–2015. Scientific Data. 4: 170095.Williamson C.E. , Saros, J.E. , Vincent, W.F. , Smol, J.P. , (2009), Lakes and reservoirs as sentinels, integrators, and regulators of climate change, Limnology and Oceanography, 54, 2273–2282.

29


Proposal Title: Improving Long-term Records of High-Resolution Satellite-derived Lake Surface Temperature for Global and Regional Climate StudiesPrincipal Investigator: Erik T. CrosmanStatement of Diversity, Inclusion, and Broader Impacts

The proposed project will result in data that will benefit society as a whole by providing improved quantitative lake temperature data that will in turn lead to improved scientific understanding and forecasting of weather, climate change, and the many important applied scientific applications that require accurate lake temperature data. In terms of NOAA’s long term climate goals to serve the general public, this proposal will advance climate intelligence and resilience through providing a data set useful for addressing (1) weather and climate extremes, (2) Climate impacts on water resources, and the (3) sustainability of marine ecosystems. The PI has carefully read NOAA’s Diversity and Inclusion Strategic Plan 2017-2019

http://www.eeo.noaa.gov/d&i/NOAA%20Diversity%20and%20Inclusion%20Strategic%20Plan.pdf and will work to further the aims and ideals set forth in that plan. In the search for the primary graduate student to assist with this study, effort will be made to ensure that this position is listed on large research opportunity emails lists and websites, to ensure equal opportunities for participation of women, persons with disabilities, and underrepresented minorities.

As part of this study, rigorous public outreach activities will be conducted. The PI will work with the VP for research office at the University of Utah to make sure recently published work is communicated via social media. An undergraduate student from the University of Utah Department of Atmospheric sciences will also be recruited during this study to conduct their senior capstone projects using the data from this study. The PI will also prepare presentations to local school groups and other educational entities. The PI has a strong working collaborative relationship with the Great Salt Lake Institute https://www.westminstercollege.edu/campus-life/centers-and-institutes/great-salt-lake-institute and outreach will be made to the GSL Institute and the Friends of the Great Salt Lake (http://www.fogsl.org/ ), using the remote sensing lake temperature climatology of this particular lake to reach this local but broad community. In addition, we will work with the College of Mines and Earth Sciences Outreach and Diversity program to present this research to junior through high school students. The outreach lectures will focus on explaining the importance of global climate change to all ecosystems and living creatures on the earth, and that through remote sensing of lakes we are effectively ‘taking the temperature’ of the earth’s surface.

30

http://www.fogsl.org/

https://www.westminstercollege.edu/campus-life/centers-and-institutes/great-salt-lake-institute

https://www.westminstercollege.edu/campus-life/centers-and-institutes/great-salt-lake-institute

http://www.eeo.noaa.gov/d&i/NOAA%20Diversity%20and%20Inclusion%20Strategic%20Plan.pdf

http://www.eeo.noaa.gov/d&i/NOAA%20Diversity%20and%20Inclusion%20Strategic%20Plan.pdf


31

Date post:	20-Aug-2018
Category:	Documents
Upload:	donhu
View:	214 times
Download:	0 times

u0198116/Aug_9_NOAA/Complet… · Web vie

Documents