Experimental Design Overview (EDO) Green Ocean Amazon …chuvaproject.cptec.inpe.br/portal/pdf/relatorios/anexo9_2013.pdf · 4 miniature unmanned aerial vehicles (UAVs) to expand

Experimental Design Overview (EDO)

Green Ocean Amazon 2014 (GOAmazon2014) Scientific Steering Committee Co-chairs: Courtney Schumacher, Texas A&M University Sue C. van den Heever, Colorado State University Members: Gilberto Fisch, IAE/CTA David Fitzjarrald, SUNY-Albany Rong Fu, UT Robert Houze, UW Richard Johnson, CSU Luiz Machado, INPE Scot Martin, Harvard Gretchen Mullendore, UND Saravanan, TAMU Maria Silva Diaz, USP Ed Zipser5, Utah Additional contributors: Paul Ciesielski (CSU), Jim Moore (NCAR/EOL), Scott Ellis (NCAR/EOL), Mark Miller (Rutgers), Tony Del Genio (NASA/GISS), and the entire DYNAMO NCAR S-PolKa team

1

Executive Summary

The critical need for Green Ocean Amazon 2014 (GOAmazon2014) arises from our currently limited understanding of deep convection over the Amazon, its susceptibility to changes in atmospheric parameters, aerosols and surface conditions, and its upscale feedbacks and atmospheric teleconnections. Furthermore, convective processes and their associated feedbacks are typically poorly represented in climate models, thereby limiting our ability to predict future weather and climate. Thus, the overarching goal of GOAmazon2014 is to fully observe the properties and life cycle of tropical convection over the Amazon in relation to environmental conditions in order to better understand and model Amazonian convection and its regional and global impacts.

The GOAmazon2014 field campaign will consist of two two-month intensive observing periods (IOPs) encompassing both the wet (1 Feb – 31 Mar) and transition (1 Sep – 31 Oct) seasons within a year-long extended observing period (EOP) during 2014. Operations will be centered on Manaus, Brazil (3°S, 60°W), which is situated in the central Amazon and experiences both pristine and highly polluted conditions. The core observations include a sounding network, geographically distributed surface sites with profiling instruments, mobile platforms, and precipitation and cloud radars. NSF-requested facilities include the NCAR S-PolKa, two ISSs, and remote-controlled boundary layer instrumentation. The NSF-component of GOAmazon2014 is designed to leverage the international collaboration already in place between the Department of Energy and Brazil that forms the basis of the current GOAmazon2014 program, a year-long campaign designed to study how aerosol and cloud life cycles are influenced by pollutant outflow from a tropical megacity.

The proposed GOAmazon2014 scientific goals require documenting the three-dimensional properties of convection ranging from shallow and non-precipitating to deep and organized, on time scales from diurnal to seasonal. They also require detailed observations of relative humidity, temperature, wind, radiation, surface fluxes, and aerosols in relation to the convection. From an observational standpoint, the diverse radar wavelengths of GOAmazon2014 will document the full spectrum of Amazonian convection. The quadrilateral sounding network and surface sites will provide data on the environments that the shallow, deep, and organized convective systems are developing in, and will be used to calculate large-scale forcing data sets for model simulations. The mobile platforms will provide observations of the spatial variability of the boundary layer and pollution plume in the vicinity of Manaus. From a temporal standpoint, the two-month long IOPs should ensure sufficient sampling in two distinct seasons to answer the posed hypotheses concerning i) the diurnal transition of convection from shallow to deep, ii) aerosol impacts on the properties of shallow and deep convection, iii) the role of the land surface on deep convection occurrence and strength in the transition season and the resulting wet season onset, and iv) the atmospheric teleconnections between the Amazon and the Atlantic basin prompted by the mesoscale convective systems observed in the wet season over the Amazon.

GOAmazon2014 will provide insight into the convective characteristics and life cycle over the Amazon, convection’s sensitivity to various environmental conditions, and the feedback of organized convective systems to larger scale processes. GOAmazon2014 will also produce an unprecedented observational data set of tropical continental convection with which to evaluate model parameterizations, thereby improving our predictions of future weather and climate.

2

1. Rationale and hypotheses The Amazon is one of the rainiest regions on earth and has a rain forest covering an area as large as

the continental United States. In general the Amazon has a highly pristine environment, although this can be dramatically altered by urban centers and biomass burning. The Amazon experiences distinct variations in shallow and deep convection on diurnal to seasonal time scales that can be attributed to any number of factors, including changes in relative humidity, lapse rates, and wind shear over a complex, forest surface. The aerosols from human activities only further muddle matters. Once Amazonian convection organizes into larger systems with significant stratiform rain components, the associated mass transports and heating potentially affect the larger scale tropical circulation, including wet season onset and atmospheric teleconnections with the Atlantic. However, these convective interactions with their environment and upscale feedbacks of organized convection are typically poorly represented in models of all scales.

Particular model biases related to these convection interactions include incorrect diurnal timing of the development and precipitation of convective storms (Mapes and Neale 2011); light (heavy) precipitation that is too frequent (infrequent) (Stephens et al. 2010); inaccuracies in the links between storm dynamics, aerosols and microphysics (Morales and Nenes 2010); biases in the convective transports of heat, moisture, and trace gases and aerosols from the boundary layer to the upper troposphere (Parazoo et al. 2011); and an inability to represent important modes of variability and convective organization such as mesoscale convective systems (Zhang and Song 2009). Such model biases significantly limit our ability to accurately predict weather and future climate. Thus, the critical need for Green Ocean Amazon 2014 (GOAmazon2014) arises from our currently limited understanding of Amazonian convection, its susceptibility to changes in thermodynamics, wind shear, aerosol concentrations, and surface conditions, its potentially significant impacts on the regional and global scale, and our inability to model all of the above interactions.

The region of Manaus, Brazil (3°S, 60°W) in the central Amazon is particularly well-suited to studying the evolution of tropical convective systems and their regional and global upscale feedbacks since it experiences a wide range of convective storm types and environmental conditions throughout the year. In addition, Manaus is a megacity of almost 3 million people, creating an isolated urban area within the otherwise pristine Amazon basin that extends for thousands of kilometers in all directions. The region downwind of Manaus varies between a very clean environment to one strongly influenced by the meandering pollution plume of the megacity. Thus, the region around Manaus is a natural laboratory, not only to observe the characteristics of the tropical continental convection, but also to study cloud-aerosol-precipitation interactions and the role of land surface processes.

The overarching goal of GOAmazon2014 is to fully observe the properties and life cycle of tropical continental convection (including shallow and deep convective cells and organized convective systems) over the Amazon in relation to thermodynamic and wind shear variations, aerosol loading, and changing surface conditions in order to better understand and model Amazonian convection and its regional and global impacts. More specifically the GOAmazon2014 hypotheses are: Hypothesis I: The diurnal evolution of Amazonian convection from shallow and non-precipitating to deep and precipitating is sensitive to a quantifiable combination of the humidity and lapse-rate in the free troposphere and cold pool variability. Hypothesis II: Under enhanced aerosol concentrations (a) the liquid water contents of shallow convective clouds will increase and the surface precipitation will decrease due to the suppression of warm rain processes; (b) the ice water contents and surface precipitation of deep convective clouds will increase as a result of convective invigoration of the updraft; (c) deep convective updrafts will be strengthened through the impacts of enhanced latent heating on buoyancy; and (d) the trends in (a)- (c) will be modulated by environmental conditions.

3

Hypothesis III: An increase in evapotranspiration and a reduction of CIN, which is dominated by plant phenology and its response to an increase in solar radiation during the dry season, contributes to an increase of continental convection during the early transition season. This increase in continental convection manifests as increases in updraft intensity and depth, enhancing diabatic heating at upper levels and initiating the reversal of the cross-equatorial flow and increasing moisture transport from the Atlantic Ocean to the Amazon. These changes provide favorable conditions for maritime convective system types, and thus pave the way to wet season onset. Hypothesis IV: The representation of diabatic heating rates associated with deep convective systems in the Amazon during the wet season is simulated incorrectly in many GCMs, thus resulting in large local rainfall biases as well as having remote climate impacts. Inaccuracies associated with cloud and land surface parameterizations, and their interaction with the convective processes, are also likely to impact the Amazon rainfall estimate bias.

These four hypotheses will be evaluated using the observations made during the GOAmazon2014 field campaign, as well as a suite of numerical models initiated, constrained and evaluated using the observational data. Further scientific details are outlined in the SPO. 2. Scientific objectives

It is envisioned that GOAmazon2014 will expedite our understanding of Amazonion convection and our efforts to improve its simulation in relation to atmospheric and biospheric variability and human activity. More specifically, the goals of GOAmazon2014 are to:

• Collect observations from the central Amazon necessary to advance our understanding of tropical

continental convection and its relationship to thermodynamic, shear, aerosol, and land-surface variations;

• Identify critical deficiencies of numerical models ranging from cloud to global scale in simulating the above tropical convective interactions;

• Provide observations to assist the broad community effort toward improving model parameterizations, especially those related to convection, land surface, and aerosol processes.

GOAmazon2014 will provide an invaluable data set from which to enhance our understanding of

tropical continental convective processes and evaluate and improve current parameterization schemes. The GOAmazon2014 data will expand on those collected previously in the Amazon, and in particular the Tropical Rainfall Measuring Mission Large-scale Biosphere-atmosphere experiment in Amazonia (TRMM-LBA), by virtue of being downwind of Manaus in the central Amazon, which experiences different synoptic forcing, different land surface characteristics, and larger extremes in aerosol concentrations than the southwest Amazon, and it will also include observations of the transition season, a vastly different environment from the wet season. 3. Experimental design

The GOAmazon2014 field campaign is currently a Department of Energy (DOE) sponsored program based in the central Amazon to study carbon, cloud, aerosol, and precipitation cycles and interactions in the vicinity of the Manaus pollution plume. Multiple Brazilian organizations and researchers are also participating in projects related to GOAmazon2014. Of particular relevance to this SPO/EDO is Brazil’s Cloud processes of tHe main precipitation systems in Brazil: a contribUtion to cloud resolVing modeling and to the GPM (Globl Precipitation Measurement)(CHUVA) campaign, which has the main objectives of studying cloud and precipitation processes throughout Brazil and validating their retrieval from GPM, an international satellite constellation with a planned launch date in early 2014. National Science Foundation (NSF) participation in GOAmazon2014 will allow for the deployment of S-PolKa, two integrated sounding systems (ISSs), and novel boundary layer measurements by Lagrangian balloons and

4

miniature unmanned aerial vehicles (UAVs) to expand the GOAmazon2014 purview to capture the properties and evolution of the full convective cloud spectrum (i.e., shallow convection, deep convection, and large stratiform rain and anvil areas) and the associated large-scale environment. As such, the GOAmazon2014 field campaign will consist of three components: • A quadrilateral sounding array (Sec. 4.1), with Rio Preto da Eva (2.7°S, 59.7°W), Itacoatiara (3.1°S,

58.4°W), Borba (4.4°S, 59.6°W), and Manacapuru (3.3°S, 60.6°W) potentially serving as the northern, eastern, southern and western vertices, respectively (Fig. 1). The sounding systems include the DOE Atmospheric Radiation Measurment (ARM) Mobile Facility (AMF1), two National Center for Atmospheric Research (NCAR) ISSs, and a CHUVA system. The AMF1 location is fixed at Manacapuru, while the other sounding system locations are still to be determined. Operational soundings taken at Manaus will also be available at 00 and 12 GMT.

• Surface and boundary layer measurements (Sec. 4.2) at the AMF1, ISSs, CHUVA K34 tower (~50 km north of Manaus), and DOE ARM Mobile Aerosol Observing System (MAOS) (Iranduba). These sites contain a mix of surface meteorology, precipitation, and radiation sensors, profiling instruments that can observe wind, temperature, and humidity structures in the planetary boundary layer, surface flux measurements, and aerosol characterization. Spatially varying lower tropospheric measurements will be made by Langrangian balloons and miniature UAVs (Sec. 4.2) and the DOE G1 (Sec. 4.3) in the vicinity of the Manaus plume.

• Radar observations (Sec. 4.4) from the NCAR S-PolKa polarimetric, dual-wavelength radar at Iranduba and the DOE vertically pointing Ka-band radar and scanning W- and Ka-band radar system at Manacapuru.

Figure 1: Proposed quadrilateral locations of sonde sites with Rio Preto da Eva (2.7°S, 59.7°W), Itacoatiara (3.1°S, 58.4°W), Borba (4.4°S, 59.6°W), and Manacapuru (3.3°S, 60.6°W) potentially serving as the northern, eastern, southern and western vertices, respectively. The DOE AMF1 will be located at Manacapuru, while the Brazilian sounding system and the two ISSs will be located at the other three vertices.

5

The main GOAmazon2014 field campaign will start on 1 January 2014 and end on 31 December 2014. It will consist of two Intensive Observing Periods (IOPs, 1 February - 31 March and 1 September - 31 October, Sec. 3.1) within an Extended Observing Period (EOP, 1 January - 31 December 2014, Sec. 3.2). As discussed in the Sec. 1 of the SPO, the region of Manaus, Brazil (3°S, 60°W) in the central Amazon experiences a wide range of convective storm types and environmental conditions throughout the year, including strong diurnal and seasonal cycles and very clean and heavily polluted atmospheric conditions. The IOPs were designed to capture the height of the wet season in February/March and the transition from dry to wet seasons in September/October and the different land surface conditions experienced during each period. TRMM Precipitation Radar (PR) observations over the Amazon between 1998-2006 during these set of months show > 200 mm/month of rain accumulation during the wet season and half that during the transition season with different spatial patterns of precipitation evident (Fig. 2, top panels). Stratiform rain fractions reach 50% and greater during the wet season and are half that in the transition season (Fig. 2, bottom panels). This decrease is in part due to a doubling of convective rain rates, so while there is less overall rain in the transition season, convective rain amounts stay roughly the same while there is significantly less stratiform rain (not shown). These TRMM PR statistics are consistent with other satellite studies showing the prevalence of large mesoscale convective systems in the wet season and more isolated and intense convection in the transition season.

Figure 2: TRMM PR V6 rainfall (top) and stratiform rain fraction (bottom) for Feb-Mar (left) and Sep-Oct (right) from 1998-2006. M indicates the location of Manaus. Figure courtesy of Courtney Schumacher and Aaron Funk.

The length of the IOPs were dictated by the need to sample a sufficient number of events to create robust statistics with which to say something quantitative about convection-environment interactions. While the diurnal cycle will be very well sampled, pinpointing variations in environmental parameters that affect the transition of convection from shallow to deep still requires a large sample size (Hypothesis I). This is even more true for determining potential aerosol impacts on convective cell characteristics since the aerosol interactions are harder to measure and are intertwined with environmental controls, making concurrent modeling efforts necessary (Hypothesis II). The average wet season onset in the central Amazon is 25 September and can be quite variable (Marengo et al. 2001), so a broad time window

6

is needed to capture it (Hypothesis III). Atmospheric teleconnections on the scale of ocean basins do not occur from a single event, but rather accumulate over many events or even a season so a long record is needed during the wet season when atmospheric teleconnections between the Amazon and Atlantic appear strongest (Hypothesis IV). The length of the IOPs also provides some protection against low rain years and helps to minimize sampling errors in atmospheric budgets (Mapes et al. 2003). Interannual variability in Amazonian rainfall has a strong link to variations of sea surface temperatures in both the Pacific and Atlantic with lower than average wet season rainfall during an El Nino and lower than average dry season rainfall when the subtropical North Atlantic is anomalously warm (Liebmann and Marengo 2001, Zeng et al. 2008). However, even in these dry periods, rain amounts do not often get dangerously low. Based upon the above sampling constraints and the synergy with the already supported DOE and Brazilian GOAmazon2014 deployments, there is little flexibility in choosing different months or years for this project. 3.1 Intensive Observing Periods (IOPs): 1 February - 31 March 2014; 1 September - 31 October 2014.

The objective of the two IOPs is to observationally document the diurnal and seasonal variations in the life cycle, organization and microphysical characteristics of Amazonian convection, as well as the vertical structure of the boundary layer and large-scale environment in which these systems develop.

This objective will be achieved through the use of radar, soundings, surface sites with profiling instruments, and mobile platoforms. Two IOPS are planned in order to represent the impacts of the wet and transition seasons on Amazonian convection. The wet season IOP will be conducted from 1 February to 31 March 2014, while the transition season IOP will be held from 1 September to 31 October 2014, giving a total of 59 and 61 days field campaign days, respectively. The two-month duration of each IOP is to ensure sufficient sampling of the convective systems and environmental variations necessary to test Hypotheses I through IV.

S-PolKa will be run continuously during the two IOPs, although the Ka-band will only likely be operated during daytime hours (i.e., to focus on the shallow to deep convective transition) to conserve the transmitter. The AMF1 vertically pointing and scanning cloud radars will also be run continuously during the two IOPs. Between IOPs, S-PolKa will remain on location and will only operate to the extent deemed necessary by S-PolKa engineers to maintain the integrity of the system. Any data collected during this time can also be used for analysis purposes. Together the radar systems will provide a comprehensive observational data set of the full convective cloud spectrum. GPS sondes will be released six times per day at each of the four proposed sounding sites for the entire period of both IOPs. Along with the two times per day operational sondes at Manaus, this will provide a total of 26 soundings taken daily in the region surrounding Manaus. This quadrilateral sounding network will provide large-scale measurements of the environment, including vertical motion and diabatic heating over the array. Wind profiler and RASS data will be collected at the ISS sites during the same time period while wind profiler, Atmospheric Emitted Radiance Interferometer (AERI), ceilometer, and lidar data will be collected at the AMF1. These lower tropospheric measurements of wind, temperature, and humidity will provide high temporal resolution (30 min) to complement the four hourly sounding launches. The DOE G1 and Lagrangian balloon and miniature UAV systems will be operated for portions of both IOPs (40 and 15-20 days, respectively), providing measurements of the spatial variability of atmospheric parameters within the boundary layer and lower troposphere, especially in the region of the Manaus pollution plume. Finally, surface measurements at the ISSs, AMF1, ARM MAOS, and CHUVA K34 (currently funded from Brazilian Agencies (FAPESP) for the first IOP only) tower will provide detailed observations of surface meteorology, rainfall, radiation, surface fluxes, and aerosols during the two IOPs.

7

3.2 Extended Observing Period (EOP): 1 January - 31 December 2014 The objective of the EOP is to understand and quantify the influence of aerosol and gaseous outflow from a tropical megacity on the carbon cycle, cloud life cycle, aerosol life cycle, and interactions between cloud, aerosols and precipitation. The EOP is the main focus of the DOE GOAmazon2014 efforts and they have already committed their AMF1 and MAOS to run continuously for all of 2014 in Manacapuru and Iranduba, respectively. During this time, the AMF1 will be launching sondes four times per day. There will also be two launches per day by the Brazilian Air Force at the operational site in Manaus. The AMF1 and MAOS sites lie downwind of Manaus so they experience the extremes of a pristine atmosphere when the Manaus pollution meanders and heavy pollution and interactions of that pollution with the natural environment when the plume regularly intersects the sites. The Brazilians plan to submit a request to extend CHUVA operations from February-March to February-December 2014, thus covering the majority of the EOP. There are no precipitation radars planned for deployment during the EOP, although there is an operational S-band radar maintained by the Amazon Protection System (SIPAM) located in Manaus. However the availability of the SIPAM data has been problematic for Brazilian researchers in the past and the quality of the data would be limited to operational needs. 4. Specific Observational Requirements Table 1: Mapping of facilities to hypotheses; E = essential, U = useful Temporal sampling I (diurnal) II (aerosol) III (onset) IV

(remote) Sounding network 6/day at 4 sites, IOPs

4/day at 1 site, EOP E E E E

Surface: Profiling instruments

ISSs and CHUVA (IOPs), AMF1 (EOP)

E E E U

Surface: Met, precip, radiation, surface fluxes

ISSs and CHUVA (IOPs), AMF1 (EOP)

E E E E

Surface: Aerosols EOP E U CMET balloons and MetSonde UAVs

15-20 flight days per IOP

U E U

DOE G1 40 flight days per IOP U E U NCAR S-PolKa radar

IOPs E E E E

DOE Ka- and W-band cloud radars

EOP E U E

4.1 Soundings

The objective of the GOAmazon2014 sounding network is to observe the diurnal and seasonal variations in the vertical structure of the large-scale environment in which Amazonian convection develops and evolves.

The sounding array proposed for use in GOAmazon2014 is shown in Fig. 1, and allows for observations of the large-scale environment both upwind and downwind of Manaus, and within and outside the typical Manaus aerosol plume that often extends from Manaus westward to Manacapuru because of the prevailing northeasterlies. As such, convective systems developing within a range of large-scale environmental and aerosol conditions may be analyzed. All of the radiosonde data will be made available to the operational centers in real time through the Global Telecommunications System (GTS) or an alternative system.

8

To estimate the sounding resources needed for GOAmazon2014, we undertook an uncertainty analysis to evaluate the errors associated with the computation of atmospheric budgets in the Amazon region. To facilitate this analysis we used high-resolution (25 km, 91 vertical levels, 6 hourly) ECMWF reanalyses from the Year of Tropical Convection (YOTC) dataset. The idea is to compute the budgets with both the high-resolution YOTC data, which provides a measure of the “true” fields, and with simulated soundings in which the winds and thermodynamic fields from the YOTC dataset are sampled at discrete virtual sounding sites. Different array shapes (triangular and quadrilateral) and sizes were considered to determine the optimal configuration for producing the most accurate budget estimates (Fig. 3). The vertices of the simulated budget arrays were located at various cities in the vicinity of Manus. The simulated soundings are then used to objectively analyze the data fields onto a regular 25 km grid. Budgets were computed on this grid using both the high-resolution data and the simulated sounding gridded analysis and then averaged over the sounding network. Computations were carried out for the two proposed IOPs (February/March to represent the wet season and September/October for the transition season) and for two years, 2008 and 2009.

Figure 3: Sounding array configuration options (polygons) and operational soundings (black circles). Sizes for the array configurations were (85 km)2, (122 km)2, and (152 km)2. The TRMM-LBA array was (95 km)2. Grid for objective analysis of simulated soundings is shown. Color shading indicates elevation over domain with scale at bottom. Figure courtesy of Paul Ciesielski.

Results from this analysis suggest that quadrilateral sonde arrays are better than triangular arrays at

resolving circulation features associated with convective structures in the Amazon region. In particular, root-mean-square (RMS) errors (i.e., the difference between high-resolution and simulated analyses) in vertically integrated heating and differences in the time-mean peak heating rates are a factor of 2-3 times smaller with a quadrilateral array compared to a triangular array. This improvement is particularly noticeable in the wet season when mean convective heating rates over the Amazon are larger.

To minimize the impact of random sampling errors on the budgets, a quadrilateral array on the order of 150-200 km2 in area appears to be optimal. Sounding arrays smaller than this size (as was the case during TRMM-LBA would have difficulty accurately capturing the large-scale forcing signal associated with convective systems in this region. Since DOE ARM (AMF1) and Brazil (CHUVA campaign) will each be providing instrumentation and resources to operate a sounding site, we are requesting two ISS systems from the NCAR deployment pool so that a quadrilateral sounding array can be constructed for the accurate computation of budgets and large-scale forcing fields for the GOAmazon2014 experiment.

9

GPS sondes will be launched six times per day at each of the four proposed quadrilateral vertices throughout both of the IOPs. Along with the twice daily Manaus operational sondes this network will provide 26 high vertical resolution profiles of the large-scale environment per day for every day of the 59 and 61 days of the wet and transitional season IOPs, respectively. Since the AMF1 is only slated to launch four sondes per day throughout the year, additional sondes will be requested from DOE to increase their frequency during the IOPs. The feasibility of working within the jungle at the non-AMF1 sites should be evaluated. Rio Preto da Eva is connected by road to Manaus and is thought to be a good location for the northern vertex. However, the east and southern vertices of the quadrilateral will need to be further evaluated.

The proposed sounding array will provide much of the data necessary to address the GOAmazon2014 Hypotheses I through IV, which rely on characterization of the vertical profiles of temperature, humidity, and wind and the diagnosed large-scale heat and moisture budgets. For example, Hypothesis I proposes that the transition from shallow to deep convection is sensitive to free-tropospheric humidity, atmospheric instability, wind shear, and boundary layer variability, including cold pools. Sounding observations are also necessary to characterize the relative humidity, vertical wind shear, and convective available potential energy (CAPE) of the large-scale environment that are expected to moderate aerosol indirect forcing for testing Hypothesis II. The relative humidity of the boundary layer, along with the boundary layer evolution and vertical wind shear are all observable quantities necessary for an investigation into Hypothesis III. Profiles of diabatic heating that can be diagnosed from the sounding network are required for testing Hypothesis III and IV. In addition, sounding data will be used to initialize, constrain, and evaluate CRM and regional model simulations relevant to all hypotheses. Together with the proposed radar, surface, and boundary layer observations, the sounding array will provide an unprecedented data set of collocated observations over a continental, tropical rain forest region of clouds, precipitation, aerosols, surface fluxes and large-scale environmental conditions. 4.2 Surface and boundary layer measurements

The objective of the GOAmazon2014 surface and boundary layer measurements is to observe diurnal and seasonal variations in land surface properties, aerosols, and low-level temperature, humidity, and wind patterns in which Amazonian convection develops and evolves. Surface sites

Three of the four sounding sites (the two ISSs and the AMF1, Fig. 1) will have a suite of surface meteorology, precipitation, and radiation sensors along with profiling instruments that can observe wind, humidity, and temperature structures in the planetary boundary layer. All measurements will be made continuously. The profiling instruments include 915 (ISSs) and 1290 (AMF1) MHz wind profilers, Radio Acoustic Sounding Systems (RASSs) for profiles of virtual temperature (ISSs), an AERI for profiles of temperature and humidity (AMF1), micropulse and Doppler lidars for cloud altitude and clear-sky winds, respectively (AMF1), and a ceilometer to measure cloud base/boundary layer heights (AMF1). The ISSs will operate during the two IOPs while the AMF1 will operate for all of 2014. The measurements at all three sites will provide high temporal resolution environmental observations with which to test Hypotheses I through III and will nicely complement the four hourly sounding launches (Sec. 4.1). The high temporal resolution will be especially important when analyzing the diurnal cycle of convection in Hypotheses I. The spatial distribution of the sites will allow characterization of clean and polluted regions when analyzing aerosol effects in Hypothesis II. The CHUVA K34 tower (which will not be collocated with the CHUVA sounding site) is located about 50 km N of Manaus and will also have surface meteorology, precipitation, and radiation sensors in addition to a Raman lidar to measure vertical profiles of moisture and aerosol scattering, a field mill to measure the strength of the electric field in the atmosphere near thunderstorms, two types of disdrometers to measure the drop-size distribution of rain, a micro rain radar, a microwave radiometer, a GPS, and soil moisture and surface flux measurements. Additional GPS, raingauge, and disdrometer measurements will made at up to three more sites (TBD). The CHUVA instrumentation is currently committed for the wet

10

season IOP (February/March), but CHUVA organizers plan to request from FAPSEP an extension of the operations to December 2014. The CHUVA focus on rain characterization will be particularly suited to helping test Hypothesis II. In addition, the surface flux measurements are important for testing the surface-atmosphere interactions in Hypothesis III and constraining the sounding budget analysis of heat and moisture profiles necessary to test Hypotheses III and IV. In addition to the AMF1 instruments already discussed, the AMF1 will mirror many of the instruments at the CHUVA K34 tower, thus contributing rain, moisture, and surface flux information to help inventory environmental and land surface variations in relation to convective cloud evolution for Hypotheses I through IV. The AMF1 also has basic chemistry and aerosol measurements to specifically address Hypothesis II. To further characterize the aerosols of the region, DOE is deploying the ARM MAOS in Iranduba, which is west and across the Rio Negro from Manaus (Fig. 1). This is also the potential site of S-PolKa. Iranduba is just downwind of Manaus and will sample the pollution plume close to the source, allowing comparisons to measurements made further downwind in Manacapuru (the plume transit normally takes 2-6 hours). The MAOS will provide invaluable information about near-surface aerosols and their relation to convection within the Manaus pollution plume with which to test Hypothesis II. Other potential sites of interest in monitoring the atmospheric chemistry of the region are an air quality monitoring station that is being installed on the Instituto Nacional de Pesquisas da Amazonia (INPA) campus in Manaus and the Amazonian Tall Tower Observatory (ATTO) that is being installed ~200 km northeast of Manaus representing observations over the pristine Amazon. Boundary layer instruments

The most novel set of boundary layer measurements will be the deployment of Controlled Meteorological (CMET) balloons (Voss et al. 2005, 2008, 2010, 2011, Riddle et al. 2006, Durant et al. 2010) and miniature MetSonde UAVs. The CMET balloons are smaller than typical rawinsondes and are equipped with altitude control, global communication via Iridium satellite, and aspirated sensors; they can fly for multiple days while performing soundings on command to probe atmospheric structure. These balloons have been flown in Brazil, along with the United States, Mexico, the Arctic, and the Antarctic. The even smaller MetSonde UAVs (250 g each) are derived from the same technology and are currently undergoing extensive testing in Antarctica. Preliminary data indicate that the MetSondes will fly for 30-40 minutes and cover approximately 20 km horizontal distance on a battery charge. The operational altitude range is expected to be approximately 1-3 km. Each instrument is equipped to measure temperature, humidity, and wind (Fig. 4). A solid-state SO2 sensor (AlphaSense Inc.) will also be integrated into the MetSondes.

Figure 4: Example CMET balloon observations near Belem, Brazil on 25 June 2011 for CHUVA (http://www.science.smith.edu/cmet/20110625_BRZ_01/flight.html).

11

The CMET balloons will generally make 3-10 soundings through the boundary layer depending on the length of the flight. Many more vertical profiles will be obtained using the MetSondes, depending on the vehicle recovery rate. Thus, the CMET balloons and MetSonde UAVs will provide spatially and vertically varying measurements of boundary layer temperature, humidity, wind, and chemistry in the vicinity of the Manaus pollution plume to complement measurements at the fixed sites discussed above, thus contributing significantly to testing Hypotheses I through III. These measurements will be particularly useful to define the boundaries of the Manaus pollution plume. Operations will occur during both IOPs, with the deployment of 12-15 balloons and five UAVs. This sampling will provide 15-20 days of observations during each IOP. Launches will be supported by in situ surface and boundary layer observations (flux tower and sodar, ASRC) at the launch site near Manaus (Iranduba is a likely location). 4.3 Aircraft The objective of the GOAmazon2014 aircraft observations is to provide lower tropospheric measurements of aerosol-cloud-precipitation interactions in the vicinity of the Manaus pollution plume.

The DOE Gulfstream G1 will be based out of the Manaus Eduardo Gomes International Airport and will fly over and around Manaus and Manacapuru. It will be deployed for 40 flight days during 15 February to 31 March and 40 flight days during 1 September to 15 October, i.e., during the wet and transition season IOPs. The G1 aircraft platform will make in situ measurements of meteorology, trace gas, aerosol, and cloud properties in the lower troposphere (its nominal operating altitude is up to 7.5 km). In brief, the droplet spectrum of clouds and precipitation will be measured by a fast cloud drop probe (F-CDP), a 2-dimensional stereo probe (2D-S), and a high-volume precipitation spectrometer (HVPS-3). A cloud imaging probe (CIP) will be used to image cloud and precipitation particles. A fast integrated mobility spectrometer (FIMS) and a passive cavity aerosol spectrometer probe (PCASP) will measure the aerosol size distribution at a time resolution of 1 Hz. An aerosol mass spectrometer (AMS) will provide measurements of the particle chemical composition. Cloud condensation nuclei (CCN) activity will be measured at multiple supersaturations (CCNC). A suite of instruments, including a nephelometer and a particle/soot absorption photometer (PSAP), will characterize aerosol optical properties. Trace gas measurements will include CO, CO2, NO, NO2, NOy, O3, and CH4. Volatile organic compounds (VOCs) will be quantified by proton-transfer mass spectrometry (PTR-MS). The meteorological information (temperature, humidity, and winds) will be measured by fast response sensors in order to estimate the mean concentration and fluxes derived. G1 observations will be used to characterize the Manaus pollution plume in time and space and will provide spatial variability in lower tropospheric measurements of environmental characteristics and aerosols with which to test Hypotheses I through III. 4.4 Radars The objective of the GOAmazon2014 radar observations is to fully characterize the properties and evolution of Amazonian convection from shallow, non-precipitating cumulus to deep cumulonimbus to organized mesoscale convective systems with large stratiform rain and anvil areas.

Advancing the understanding of the life cycle and properties of Amazonian convective systems, which is integral to all four GOAmazon2014 hypotheses, requires observations of the full convective spectrum. Dynamics of the Madden-Julian Oscillation (DYNAMO) had a similar requirement for studying the initiation of the Madden-Julian Oscillation in the Indian Ocean and much was learned during that deployment that will be utilized in GOAmazon. As in DYNAMO, observations of as many variables as possible (e.g., areas and heights of radar echoes, types and concentrations of hydrometeors, air motions internal to clouds, separation of precipitating and non-precipitating components of clouds, subdivision of precipitating clouds into convective and stratiform components, structure and organization of cloud ensembles, and propagation characteristics) with multiple wavelengths (i.e., mm- and cm-wavelength) are essential to the GOAmazon2014 radar objective. The radar observations must also encompass statistics of the convective cloud population as it transitions from shallow to deep diurnally and from deep to

12

organized seasonally. Surface-based radars are the only instruments that can routinely observe these features of the cloud population and resolve them in three-dimensions over a large spatial domain and over a reasonably long time.

Since an individual radar wavelength can only observe a subset of the convective spectrum over the Amazon, GOAmazon2014 will employ a diversity of wavelengths. In particular, DOE will deploy a number of mm-wavelength radars (i.e., a vertically pointing Ka-band system and a scanning Ka- and W-band system) more sensitive to cloud particles and small hydrometeors (but more susceptible to attenuation) than cm-wavelength radars. These radars will be deployed at the AMF1 in Manacaparu, which is about 70 km west of Manaus (Fig. 1), for all of 2014. NCAR will deploy S-PolKa, an S-band (10-cm) radar that is less prone to attenuation and better suited to observing precipitation-size hydrometeors than mm-wavelength radars, just west of Manaus (Fig. 1) for two months during the wet season and two months during the transition season. S-PolKa also has a beam-matched Ka-band (1-cm) allowing more accurate dual-wavelength retrievals. In addition, CHUVA plans to submit a proposal to deploy its X-band (3-cm) polarimetric radar between the AMF1 and CHUVA K34 tower from February-December 2014. Although X-band rapidly attenuates in heavy rain and convection, this deployment would provide some polarimetric precipitation radar coverage for the majority of the EOP. An operational S-band radar operated by SIPAM is located at Manaus, but the quality and availability of the data remain unclear. Brazilian researchers have no ability to change the scan strategy and have significant difficulties obtaining the data. However, we are continuing to pursue this as a possible surveillance radar that would provide more flexibility in S-PolKa’s scan strategy and provide areal precipitation observations during the rest of the EOP. If the NSF-component of GOAmazon2014 is supported, we will visit SIPAM to forge a scientific partnership with their organization. The main capabilities of the radars are listed below. a) NCAR S-band polarimetric radar (S-PolKa) Ka-band: • Document the 3D reflectivity and Doppler velocity structures of small non-precipitating cumulus

clouds (Nuijens et al. 2009) and their development into small cumulonimbus out to a range of 50 km S-band: • Observe 3D reflectivity, Doppler velocity, and polarimetric structures associated with the full

precipitating cloud spectrum and its evolution from small to deeper and wider convective systems out to a range of 150 km

Combined Ka- and S-band: • Determine differential attenuation between two wavelengths for low-level moisture retrievals b) DOE Ka-band ARM zenith radar (KAZR) • Document vertical profiles of reflectivity, Doppler velocity, and turbulence associated with clouds

and light rain c) DOE W- and Ka-band scanning ARM cloud radars (SACRs) • Observe the 3D reflectivity and Doppler velocity structures in nonprecipitating clouds and lightly

raining clouds out to a range of 30-50 km As was successfully done in DYNAMO, a variety of scan strategies and analysis techniques will be used to optimize the information gained from each wavelength. Scan design

All of the radars will be operated continuously except for the Ka-band of S-PolKa, which will likely only be operated during the day to conserve the transmitter. S-PolKa will have a mix of attended and remote operation, while the DOE radars will be run unmanned. The DOE vertically pointing radar does not require a scan strategy, while the scan strategy of the remaining radars will be coordinated to optimize observations. Example scan strategies are as follows and are based on what was done in DYNAMO. The

13

AMF1 scanning cloud radars will collect a combination of vertical profiles, plane position indicator (PPI) scans, and range-height indicator (RHI) scans on a 30-min cycle to document the characteristics of shallow non-precipitating convection and the weakly precipitating portions and anvil clouds of deeper convective systems. S-PolKa will perform a combination of PPI and RHI scans on a 15-min cycle to document the precipitating portions of the convection and to obtain high-resolution polarimetric microphysical observations and dual-wavelength LWC and humidity information (Figs. 5 and 6). Half of the 15-min cycle will be dedicated to a full volume scan of 12 elevation angles assuming a 10°/s antenna speed, the other half of the cycle will be dedicated to approximately 38 RHIs set 1-2° apart assuming a 5°/s antenna speed. A long-range surveillance scan may also be included and would be dependent on data access to the SIPAM operational radar. The S-PolKa full volume scan will provide the broad horizontal coverage necessary for rain maps, 3D climatologies of storm structure, dual-wavelength moisture retrievals, and divergence profile calculations, while the RHI sectors will provide high vertical resolution with which to retrieve physically meaningful microphysical information. The RHI sectors will potentially be split in a bowtie pattern with half of the bowtie facing downwind toward Manacaparu and the AMF1 and the other half facing upwind of Manaus into (typically) non-polluted conditions. Both the PPI and RHI scans performed by S-PolKa will provide storm structure context for the AMF1 cloud radars.

Figure 5: S-PolKa S-band radar data for 0220-0232 UTC 12 November 2011. Upper panel shows PPIs of reflectivity (left) and differential reflectivity (right) at 0.5 deg elevation. Bottom panel shows reflectivity (left) and radial velocity (middle) and particle type (right) along the yellow lines in the PPIs. From the DYNAMO S-PolKa science summary (http://catalog1.eol.ucar.edu/dynamo/).

14

Radar data analysis The analysis of the data obtained by the GOAmazon2014 radars will be based on known technology

and will build upon the successful integration of the different wavelengths and techniques in DYNAMO. The following is a list of radar retrievals that will be used for GOAmazon2014 hypothesis testing. Figures 6 and 7 show example S-PolKa results for some of these retrievals on 12 November 2011. • Obtain high-quality polarimetric-based maps of rain rate within ~150 km of the radar (Brandes et al.

2002) • Classify hydrometeor types, including distinguishing between dry snow, wet aggregates, dry

aggregated and/or graupel, and rain (Vivekanandan et al. 1999a) • Diagnose convective and stratiform rain fractions (Churchill and Houze 1984; Steiner et al. 1995;

Yuter and Houze 1997) • Provide estimates of latent heating profiles (adapting techniques used by Schumacher et al. 2004 and

Shige et al. 2004, 2007) and comparing these with sounding-based heating profiles (Tao et al. 2006, 2007a)

• Calculate statistical information on the divergence profiles within the storms from single-Doppler measurements (Mapes and Lin 2005)

• Quantify cold pool initiations (Lima and Wilson 2008) • Obtain estimates of mean vertical profiles of lower troposphere humidity by differential atmospheric

attenuation measurements (Ellis and Vivekanandan 2010) • Obtain estimates of total cloud LWC using the differential liquid water attenuation measurements

(Vivekanandan et al. 1999b, Ellis and Vivekanandan 2011) - the LWC estimates include the contributions from cloud droplets and drizzle/rain

• Retrieve high-resolution profiles of cloud droplet microphysics and cloud radiative heating in conjunction with other measurements (McFarlane et al. 2007)

• Provide a dataset to retrieve high-resolution 3D cloud radiative heating in conjunction with other measurements (McFarlane et al. 2007)

Figure 6: Dual-wavelength retrievals of relative humidity from S-PolKa on 12 November 2011 compared to the Gan Island soundings during DYNAMO. Figure courtesy of Scott Ellis and Scott Powell.

15

Figure 7: Upper panel: Convective, stratiform and total rain seen by the S-PolKa S-band radar by rain amount and rain area for 12 November 2011. Lower panel: Frequency of occurrence of convective and stratiform echo tops at different reflectivity thresholds as seen by the S-PolKa S-band radar on 12 November 2011. From the DYNAMO S-PolKa science summary (http://catalog1.eol.ucar.edu/dynamo/).

The GOAmazon2014 radar observations will be used to construct cloud and precipitation statistics at different stages of a convective system’s evolution in conjunction with observations of the larger scale environment provided by boundary layer measurements, soundings, satellite data, and reanalyses. GOAmazon2014 radar-derived statistics will be compared to longer satellite records (such as from TRMM and CloudSat) in order to assess their representativeness and can potentially be used in algorithm validation (e.g., for GPM and Megha-Tropiques). GOAmazon2014 radar statistics will be used to validate and calibrate the representation of convection in numerical models, upon which tests of GOAmazon2014 hypotheses must ultimately rest.

16

Contribution of radar observations to the GOAmazon2014 hypotheses Hypotheses I and II require detailed observations of both shallow and deep convective cells in order to analyze the diurnal transition between these two convective cloud types and how their structural properties are impacted by environmental variations and aerosols. Radar measurements and retrievals of convective cell horizontal size and height, storm initiation mechanisms, the separation between precipitating and non-precipitating regions, relative humidity, LWC, hydrometeor types and concentrations, and surface rainfall will be essential to testing the first two hypotheses. Hypotheses III and IV require detailed observations of deep convective cells and stratiform rain and anvil regions in order to analyze the seasonal transition between these convective cloud types, how their structural properties are affected by land-surface variations, and how they impact the larger scale tropical circulation through diabatic heating. Radar measurements and retrievals of echo-top heights, surface rainfall separated into convective and stratiform components, divergence profiles, latent and radiative heating profiles, and overall structure and organization of cloud ensembles will be essential to testing the third and fourth hypotheses. 5. Project management and operational support The GOAmazon2014 Science Steering Committee (see cover page) provides overall scientific guidance to GOAmazon2014 planning and execution. The Principal Investigator (PI), Courtney Schumacher, and Co-PI, Sue van den Heever, are responsible for the overall planning, coordination, and operation of the field experiment. They will work to coordinate activities with the broader GOAmazon2014 science team (including DOE and Brazilian participants), the GOAmazon2014 Project Office, the DOE ARM Project Office, and NSF facility managers at the NCAR Earth Observing Laboratory (EOL). Data management activities will be provided by NCAR EOL as described in the following section. GOAmazon2014 will request EOL to provide assistance and advice in the development of the GOAmazon2014 Project Office. The office will consist of the GOAmazon2014 PIs, staff from NCAR/EOL, and NSF program officers. The GOAmazon2014 Project Office will be tasked to coordinate all ground-based facilities to be used in the field campaign, project communications, and in-field and post-field data management. The Project Office will participate in planning discussions and meetings as required and assist with the preparation of the GOAmazon2014 Operations and Data Management Plans. Critical tasks in the period prior to the field deployment will include site visit and logistics planning and coordination for surface-based instrument sites for several participating facilities. The Project Office will also assist with pre-experiment arrangements to ensure that relevant field measurements during GOAmazon2014 are successfully transmitted onto the GTS and used by operational centers. During the field deployment, the Project Office will assist with conducting regular planning meetings, the dissemination of critical project planning information, and the preparation and implementation of a GOAmazon2014 Field Catalog to use for daily planning and decision making. 6. Data policy and management The GOAmazon2014 data policy is in compliance with the World Meteorological Organization (WMO) Resolution 40 on the policy and practice for the exchange of meteorological and related data and products including guidelines on relationships in commercial meteorological activities: "As a fundamental principle of the World Meteorological Organization (WMO), and in consonance with the expanding requirements for its scientific and technical expertise, the WMO commits itself to broadening and enhancing the free and unrestricted international exchange of meteorological and related data and products." Additional GOAmazon2014 data policy requires: • A GOAmazon2014 Data (field observations and associated satellite data, reanalyses, and model

output) Archive Center (DDAC) will be established and maintained by NCAR EOL. • A real-time web-based Field Catalog will be implemented by EOL to assist the planning and field

operation with an overview of the missions carried out during the field campaign. All participants to

17

the GOAmazon2014 field campaign are required to communicate with EOL on a daily basis to report status of their real-time data collection and instruments, which will be included in the Field Catalog. Near real-time radar images and Skew-T plots based on atmospheric sounding observations will be provided in the Field Catalog.

• Within six months following the end of the field campaign, all data shall be promptly provided by GOAmazon2014 investigators responsible for data acquisition to other GOAmazon2014 investigators upon request and notification of the intent of data use.

• All GOAmazon2014 investigators participating in the field campaign are required to submit their field data to the DDAC no later than six months following the end of the field campaign.

• During the first 12 months following the end of the field campaign, all GOAmazon2014 data will be accessible only to GOAmazon2014 investigators to facilitate inter-comparison, quality control checks and inter-calibrations, as well as an integrated interpretation of the combined data set. No public release of the data (sharing with non-GOAmazon2014 colleagues, conference presentations, publications, commercial and media use, etc.) is allowed without the permission of the GOAmazon2014 PIs who are responsible for collecting the data.

• Quality control procedures should be carried out by GOAmazon2014 investigators within 12 months following the end of the field campaign, unless unforeseeable issues emerge. After that, GOAmazon2014 field data will be made available to the broader scientific community. Any remaining data quality issues should be made clear in the data documentation files. Improving GOAmazon2014 data quality will be a continuous effort. The suitability of the released data for scientific investigations and publications should be decided at the discretion of the GOAmazon2014 investigators responsible for field data collection and quality control and data users.

• The authorship decision for publications resulting from using GOAmazon2014 data should follow the ethic rules of the journals and professional organizations (e.g., AMS, AGU). GOAmazon2014 investigators responsible for field data collection are encouraged to make contributions to data analysis and writing of manuscripts, in addition to providing the data, to be co-authors of publications using GOAmazon2014 data.

• The following acknowledgements are suggested to be included in all publications using GOAmazon2014 data: The xxxx data were collected as part of GOAmazon2014, which was sponsored by NSF and DOE. The involvement of the NSF-sponsored NCAR EOL is acknowledged. [The acquisition of the xxx data was carried out by YYYY under the support by wwww (if YYYY is not a co-author)]. The data are archived at the GOAmazon2014 Data Archive Center maintained by NCAR EOL.

18

7. References Brandes, E. A., G. Zhang, and J. Vivekanandan, 2002: Experiments in rainfall estimation with a

polarimetric radar in a subtropical environment. J. Appl. Meteor., 41, 674–685. Churchill, D. D., and R. A. Houze, Jr., 1984: Development and structure of winter monsoon cloud

clusters on 10 December 1978. J. Atmos. Sci., 41, 933–960. Durant, A., P. B. Voss, M. Watson, T. Roberts, H. Thomas, F. Prata, J. Sutton, T. Mather, M. Witt, and

M. Patrick, 2010: Real-time in situ measurements of volcanic plume physio-chemical properties using Controlled METerological balloons. General Assembly, Vienna, Austra, European Geophysical Union.

Ellis, S. M., and J. Vivekanandan, 2010: Water vapor estimates using simultaneous dual- wavelength radar observations. Radio Sci., 45, RS5002.

Ellis, S. M., and J. Vivekanandan, 2011: Liquid water content estimates using simultaneous S and Ka band radar measurements. Radio Sci., 46, RS2021.

Liebmann, B., and J.A. Marengo, 2001: Interannual variability of the rainy season and rainfall in the Brazilian Amazon basin. J. Clim., 14, 4308-4318.

Lima, M. A., and J. W. Wilson, 2008: Convective storm initiation in a moist tropical environment. Mon. Wea. Rev., 136, 1847-1864.

Mapes, B. E., and J. L. Lin, 2005: Doppler radar observations of mesoscale wind divergence in regions of tropical convection. Mon. Wea. Rev., 133, 1808-1824.

Mapes, B.E., and R. B. Neale, 2011: Parameterizing Convective Organization to Escape the Entrainment Dilemma, J. Adv. Model. Earth Syst., 3, M06004, doi:10.1029/2011MS000042.

Mapes, B. E., P. E. Ciesielski, and R. H. Johnson, 2003: Sampling errors in Marengo, J. A., B. Liebmann, V. E. Kousky, N. P. Filizola, and I. C. Wainer, 2001: Onset and end of the rainy season in the Brazilian Amazon Basin. J. Climate, 14, 833-852.

McFarlane, S. A., J. H. Mather, and T. P. Ackerman, 2007: Analysis of tropical radiative heating profiles: A comparison of models and observations. J. Geophys. Res., 112, D14218.

Morales, R. and Nenes, A., 2010: Characteristic updrafts for computing distribution-averaged cloud droplet number, autoconversion rate and effective radius, J. Geophys. Res., 115, D18220, doi:10.1029/2009JD013233.

Nuijens, L., B. Stevens, and A. P. Siebesma, 2009: The environment of precipitating shallow cumulus convection. J. Atmos. Sci., 66, 1962-1979.

Parazoo, N. C., A. S. Denning, J. A. Berry, A. Wolf, D. A. Randall, S. R. Kawa, O. Pauluis, and S. C. Doney, 2001: Moist synoptic transport of CO2 along the mid-latitude storm track. Geophys. Res. Lett., 38, L09804, doi:10.1029/2011GL047238.

Riddle, E. E., P. B. Voss, A. Stohl, D. Holcomb, D. Maczka, K. Washburn, and R. W. Talbot, 2006: Trajectory model validation using newly developed altitude-controlled balloons during the International Consortium for Atmospheric Research on Transport and Transformations 2004 campaign. J. Geophys. Res.-Atmos., 111, 13.

Schumacher, C., R. A. Houze, and I. Kraucunas, 2004: The tropical dynamical response to latent heating estimates derived from the TRMM precipitation radar. J. Atmos. Sci., 61, 1341-1358.

Shige, S., Y. N. Takayabu, W.-K. Tao, and C. L. Shie, 2007: Spectral retrieval of latent heating profiles from TRMM PR data. Part II: Algorithm improvement and heating estimates ver tropical ocean regions. J. Appl. Meteor., 46, 1098-1124.

Shige, S., Y. N. Takayabu, W.-K. Tao, and D. E. Johnson, 2004: Spectral retrieval of latent heating profiles from TRMM PR data. Part I: Development of a model-based algorithm. J. Appl. Meteor., 43, 1095-1113.

Steiner, M., R. A. Houze, Jr., and S. E. Yuter, 1995: Climatological characterization of three-dimensional storm structure from operational radar and rain gauge data. J. Appl. Meteor., 34, 1978-2007.

Stephens, G. L., T. L'Ecuyer, R. Forbes, A. Gettlemen, J.-C. Golaz, A. Bodas-Salcedo, K. Suzuki, P.

19

Gabriel, and J. Haynes Dreary state of precipitation in global models, 2010: J. Geophys.Res., 115, D24211, doi:10.1029/2010JD014532.

Tao, W.-K., E. Smith, R. Adler, Z. Haddad, A. Hou, T. Iguchi, R. Kakar, T.N. Krishnamurti, C. Kummerow, S. Lang, R. Meneghini, N. Nakamura, T. Nakazawa, K. Okamoto, W. Olson, S. Satoh, S. Shige, J. Simpson, Y. Takayabu, G. Tripoli, and S. Yang, 2006: Retrieval of latent heating from TRMM measurements. Bull. Amer. Meteor. Soc., 87, 1555-1572.

Tao, W.-K., R. A. Houze, Jr., and E. A. Smith, 2007a: The fourth TRMM latent heating workshop. Bull. Amer. Meteor. Soc., 88, 1255-1259.

Vivekanandan, J., B. E. Martner, M. K. Politovich, and G. Zhang, 1999b: Retrieval of atmospheric liquid and ice characteristics using dual-wavelength radar observations. IEEE Trans. on Geos. and R. Sens.. 37, 2325-2334.

Vivekanandan, J., D. S. Zrnic, S. M. Ellis, R. Oye, A. V. Ryzhkov, and J. Straka, 1999a: Cloud microphysics retrieval using S-band dual-polarization radar measurements. Bull. Amer. Meteor. Soc., 80, 381-388.

Voss, P. B., and Coauthors, 2010: Long-range pollution transport during the MILAGRO-2006 campaign: a case study of a major Mexico City outflow event using free-floating altitude-controlled balloons. Atmos. Chem. and Phys., 10, 7137-7159.

Voss, P. B., E. E. Riddle, and M. S. Smith, 2005: Altitude control of long- duration balloons, J. Aircr., 42, 478–482.

Voss, P.B., “System and method for altitude control", U.S. Pat. No. 7,469,857B2, Issued Dec. 30, 2008. Voss, P.B., L. R. Hole, A. Mentzoni, E. F. Helbling, H. G. Johnston, and T. J. Roberts, 2011:

Controllable Meteorological Balloons for Arctic Research, 11th AIAA Aviation Technology, Integration, and Operations (ATIO) Conference, including the AIAA Balloon Systems Conference and 19th AIAA Lighter-Than-Air Technology Conference, Virginia Beach, VA.

Yuter, S. E., and R. A. Houze, Jr., 1997: Measurements of raindrop size distributions over the Pacific warm pool and implications for Z-R relations. J. Appl. Meteor., 36, 847-867.

Zeng, N., J. H. Yoon, J. A. Marengo, A. Subramaniam, C. A. Nobre, A. Mariotti, and J. D. Neelin, 2008: Causes and impacts of the 2005 Amazon drought. Envir. Res. Letters, 3, 9.

Zhang, G. J., and X. Song, 2009: Interaction of deep and shallow convection is key to Madden-Julian Oscillation simulation, Geophys. Res. Lett., 36, L09708, doi:10.1029/2009GL037340.

20

Section I: Facilities, Equipment and Other Resources

Field facility request for GOAmazon2014 Facility Period Sponsor Request

status Cost PI

S-PolKa radar IOPs NSF-DPL TBS $1.6M C. Schumacher (Texas A&M) 2 ISSs and supplemental sondes for Brazilian system

IOPs NSF-DPL TBS $1.2M S. van den Heever (CSU)

CMET balloons and MetSonde UAVs

IOPs NSF-ATM TBS $270K D. Fitzjarrald (SUNY-Albany) P. Voss (Smith College)

Project management EOP NSF TBS $200K J. Moore (NCAR/EOL) Data management EOP + 1

year NSF TBS $400K S. Williams (NCAR/EOL)

AMF1, MAOS EOP G1 IOPs

DOE

Funded $3.6M S. Martin (Harvard)

Sounding system IOPs Surface sites 1st IOP*

Funded

X-Pol radar EOP

Brazil/ FAPESP

TBS

N/A L. Machado (INPE)

Summary NSF-DPL: $2.8M

NSF-ATM: $270K NSF: $600K DOE: $3.6M Total: $7.27M

NSF-DPL: NSF deployment pool TBS: To be sumitted *FAPESP request to extend surface site deployments through the end of the EOP is TBS.

21

Section J: Special Information and Supplementary Documentation Characterization of the properties and life cycle of Amazonian convection during GOAmazon2014 PI: Courtney Schumacher, Department of Atmospheric Sciences, Texas A&M University ([email protected]) Anticipated funding agencies: NSF Deployment Pool and NSF-ATM Duration and starting date: 3 years, 1 July 2013 Estimated budget: Preparation and field phase, $1.8M (NSF deployment pool: $1.6M and NSF-ATM: $200K) (Year 1); data analysis, $150K (Year 2), data analysis, $160K (Year 3)

The Amazon is one of the rainiest regions on earth and experiences distinct variations in shallow and deep convection on diurnal to seasonal time scales that can be attributed to any number of factors, including changes in relative humidity, lapse rates, and wind shear over a complex, forest surface. The aerosols from human activities only further complicate matters. Once Amazonian convection organizes into larger systems with significant stratiform rain components, the associated mass transports and heating potentially affect the larger scale tropical circulation, including wet season onset and atmospheric teleconnections with the Atlantic. However, these convective interactions with their environment and upscale feedbacks of organized convection are typically poorly represented in models of all scales.

Advancing the understanding of the life cycle and properties of Amazonian convective systems requires observations of the full convective spectrum. The radar observations must also encompass statistics of the convective cloud population as it transitions from shallow to deep diurnally and from deep to organized seasonally. Thus, this proposal will focus on the deployment and data analysis of the NCAR polarimetric S- and Ka-band Doppler radar (S-PolKa) during GOAmazon2014. S-PolKa will be located near Manaus in the central Amazon for two 2-month intensive observing periods during 2014: February/March (wet season) and September/October (transition season) capturing a range of convective and environmental variations.

S-PolKa observations and products are considered essential to testing all of the GOAmazon2014 hypotheses and will be the focus of our data analysis. Hypotheses I and II require detailed observations of both shallow and deep convective cells in order to analyze the diurnal transition between these two convective cloud types and how their structural properties are impacted by environmental variations and aerosols. Radar measurements and retrievals of convective cell horizontal size and height, storm initiation mechanisms, the separation between precipitating and non-precipitating regions, relative humidity, LWC, hydrometeor types and concentrations, and surface rainfall will be essential to testing the first two hypotheses. Hypotheses III and IV require detailed observations of deep convective cells and stratiform rain and anvil regions in order to analyze the seasonal transition between these convective cloud types, how their structural properties are affected by land-surface variations, and how they impact the larger scale tropical circulation through diabatic heating. Radar measurements and retrievals of echo-top heights, surface rainfall separated into convective and stratiform components, divergence profiles, latent and radiative heating profiles, and overall structure and organization of cloud ensembles will be essential to testing the third and fourth hypotheses.

22

Latent Heating, Convective Invigoration by Aerosols and Modulation of Aerosol Indirect Effects Derived from GOAmazon2014 Sounding Data PI: Susan C. van den Heever, Colorado State University ([email protected]) Anticipated funding agency: National Science Foundation, NSF-ATM Start date: 1 July 2013, 3 year award Estimated budget: Preparation and field phase, $1400K* (Year 1); data analysis, scientific research $190K, Year 2; data analysis, scientific research, $195K, Year 3. This proposal is directed toward the planning and implementation of a sounding network for GOAmazon and subsequent quality control and analysis of the sounding data. The scientific objective of this network is to provide sounding-based estimates of vertical profiles of divergence and diabatic heating for two IOPs (a wet season from February-March 2014 and a transition season from September-October 2014) over the Amazon basin near Manaus, Brazil. The measurements are intended to complement aerosol measurements and independent observations of the evolving cloud and precipitation fields from radars, satellites and the DOE AMF1 instruments. The sounding data set will also be used (1) to investigate the modulation of aerosol indirect effects by vertical wind shear, CAPE, lower tropospheric stability and relative humidity; (2) to examine the latent heating fields and associated convective development throughout the convective life cycle within clean and polluted conditions; and (3) to construct large-scale forcing fields for initializing and evaluating CRM and regional model simulations and to assist in improving model parameterization schemes. The proposed sounding network consists of four sites in a quadrilateral shaped array centered near Manaus (3.1°S, 61.0°W). The recommended sites include Rio Preto da Eva (2.7°S, 59.7°W), Itacoatiara (3.1°S, 58.4°W), Borba (4.4°S, 59.6°W), and Manacapuru (3.3°S, 60.6°W). Integrated Sounding Systems or ISSs (including surface instrumentation, a GAUS sonde system, a 915-MHz clear-air wind profiler and a Radio-Acoustic Sounding System or RASS) are requested for two of the sites (locations yet to determined). Soundings for the other sites will be provided by DOE ARM (at Manacapura) and Brazil (location yet to be determined). During the two IOPs, lasting 59 and 61 days, respectively, there will be 6 sounding launches per site per day designed to sample the diurnal evolution and life cycle of convection within this region. During the field program, our CSU team will monitor the sounding data quality as well as its transmission to operational centers via the GTS. Following the experiment, we will carry out quality control of the sounding data working with other groups from NCAR EOL, DOE and Brazil, paying particular attention to possible humidity errors. Quality controlled sounding data will be released to the GOAmazon community in 2015 and efforts will begin on diagnostic studies of convection addressing Hypotheses 1 and 3 using data from the sounding network and other observations taken during GoAmazon, as well as numerical simulations. * Includes $1200K for NSF Deployment Pool sounding costs for two ISSs.

23

Innovative boundary layer observations to advance understanding of mixing processes and trajectories downwind of Manaus PIs: David Fitzjarrald, ASRC, University at Albany, SUNY;Paul Voss, Smith College, Anticipated funding agency: NSF-ATM Duration: 3 years, starting date 1 July 2013 Estimated budget: Deployment costs: $228,591 (if both IOPs $270,000); Data analysis/modeling (years 2 & 3 total: $184,000); Average cost/year: $150,700. Science Summary: Near the great rivers of the Amazon Basin breezes perturb the boundary layer (BL) flow. Mesoscale circulations that arise in response to such breezes as well as roughness contrasts also alter precipitation patterns. These local circulations also complicate pollutant transport downwind of cities. Transient pressure gradients of the same order of magnitude (≈0.5hPa/100km) can also be provoked near deep convection, further influencing real parcel trajectories in the BL. Currently modeled Lagrangian tracer work suffers from a lack of recent field direct trajectory observations, particularly atmospheric trace constituents. Observations and instruments: Controlled Meteorological (CMET) balloons and miniature MetSonde UAVs are to be deployed. The quasi-Lagrangian CMET balloons can fly for multiple days, performing soundings on command. The even smaller MetSonde UAVs, derived from the same technology, fly for 30-40 minutes (≈ 20 km) with altitude range 1-3 km For GOAmazon 12-15 CMET balloons and 3-4 MetSondes would profile Θ, q, and the wind vector. Each CMET balloon will generally make 3-10 soundings through the boundary layer. Many more vertical profiles will be obtained using the MetSondes, depending on the vehicle recovery rate. A solid-state SO2 sensor on MetSondes will be used to probe into the edges of the Manaus plume. The MetSondes will perform repeated vertical and horizontal soundings to determine the cross-section of the Manaus plume. Analysis: Field observations are to be followed by an in-depth study of boundary layer processes operating during the intensive study period. Our efforts aim to improve understanding of the importance of local circulations on trajectories originating in Manaus (Lagrangian tracer effort) using the CMET, These data can trace downwind variation of these parameters and will help to estimating detrainment effects at CBL top as well as at plume lateral boundaries. With our Brazilian colleagues mesoscale model simulations are planned during the analysis phase. This work will also examine the importance of a correct boundary layer description of plume dispersion in the presence of vertical wind direction shear and entrainment. Under what circumstances is the relatively simple ‘box-model’ description adequate to describe the evolution of reactive species downstream of an urban area? Intensive measurements of the vertical and horizontal profiles of selected atmospheric constituents can open the way to improve current models. Contribution to Project Goals: Our aims are to: a) relate the along-stream variation of these variables within and just above the atmospheric boundary layer (ABL) to the underlying surface and to the ambient cloud environment; b) better understand the role of river/land breeze effects on boundary layer wind shear and consequent air mass advection; and c) observe profiles of Θ, q, selected trace gases, and wind vector to assess the degree to which current ABL modules can describe entrainment and lateral mixing.

24

Ensemble forecasts of convective variability during GOAmazon2014 PI: Gretchen Mullendore, University of North Dakota (UND), [email protected] Anticipated funding agency: NSF-ATM Duration and starting date: 3 years, 15 August 2013 Estimated Budget: Preparation, field operations, and undergraduate education, 222K1,2 (Year 1); data analysis and undergraduate education, 92K2 (Year 2); data analysis, 80K (Year 3) Operational Ensemble This proposal will focus on the simulations of the GOAmazon for both wet season and transition season IOPs. In collaboration with Brazilian researchers, a minimum of five realizations per day of convective activity will be forecasted to serve as a small ensemble. Prior to the initial deployment, different model physics and initialization schemes will be tested and compared to previous Amazonia field observations (e.g. TRMM-LBA) to best capture observed convection and give realistic ensemble spread. During the campaign, key forecast fields (e.g. temperature and wind profile, convective timing) will be verified daily to judge best-performing members. The ensemble members will be used to aid in forecasting for mobile platforms. In addition, analysis and improvement of simulated tropical continental convection is a major goal of GOAmazon; forecasted fields from these simulations will be available to researchers post-campaign via the campaign catalog and will serve as support to hypothesis testing, particularly hypotheses 1, 2 and 3. Undergraduate Education In Fall 2013, the existing junior/senior UND undergraduate course that covers atmospheric modeling will have an extra focus on ensemble modeling. Students in the course (~15) will participate in the testing of ensemble members proposed above. Two of the top students will be hired during Spring 2014 to run one of the ensemble members on the UND Atmospheric Science undergraduate computer cluster. Those students will be funded to visit the campaign over an extended spring break (2 weeks) and encouraged to continue research as part of their senior capstone projects, to be presented at the AMS Student Conference in Fall 2015. In collaboration with the UND Office of International Programs, more than two students will be encouraged to participate in the spring break trip, but only two will be funded as undergraduate researchers. Data Analysis Data analysis includes visualization of forecasted fields for the campaign catalog. Data analysis in years 2 and 3 will focus on the impacts of convective variability on deep convective transport (hypotheses 1 and 3). Although no direct upper-level chemical measurements will be made during GOAmazon, methodology developed at UND has shown that upper-level radar scans can be used as a proxy for determining vertical profiles of mass detrainment in deep convection. This same technique will be employed and compared with other observed parameters, including convective mode and variability due to shallow convection, to better understand which factors most strongly impact depth of deep convective transport. 1Includes 98K for additional computer equipment and on-site forecasting support for GOAmazon ensemble simulations; 2Includes 72K for undergraduate education component, over 2 years

25

Analysis and modeling of GOAmazon2014 Field Observations PI: Rong Fu, The University of Texas at Austin Expected Funding Agency: NSF-ATM Duration and Starting Date: 3 years from September 2014 Estimated Budget per year: $136K/Y1, $141/Y2, $146/Y3 The main goal of our project is to test hypothesis III: An increase in evapotranspiration and a reduction of CINE, which is dominated by plant phenology and its response to an increase in solar radiation during the dry season, contributes to an increase of continental convection during the early transition season. This increase in continental convection manifests as increases in updraft intensity and depth, enhancing diabatic heating at upper levels and initiating the reversal of the cross-equatorial flow and increasing moisture transport from the Atlantic Ocean to the Amazon. These changes provide favorable conditions for maritime convective system types, and thus pave the way to wet season onset. Our focus is on role of vegetation and land surface, we will collaborate with other efforts that focus on transition from shallow to deep convection and aerosols indirect effect. The proposed activities include: a) Determining the relative contribution of an increasing ET and large-scale moisture transport

to changes of humidity in the atmospheric boundary layer and CIN. We will analyze surface solar, sensible, latent and CO2 fluxes measured at K34 tower, AMF1, NCAR ISSs, radionsonde profiles data. Strong correlation between CO2 and ET fluxes at canopy top suggests that ET and CO2 fluxes are dominated by plants photosynthesis and transpiration in responding to change of solar radiation. We can also estimate the influence of ET vs. horizontal advection and entrainment at the top of the boundary layer (BL) on humidity change in the BL through joint use of flux tower data, the radiosound data and AMF1. The analysis aims to determine how vegetation responds to changes of the solar radiation influences surface sensible and latent flux, which in turn, influence daytime BL height and humidity and CIN? What is the relative contribution between ET and entrainment at the top of the boundary layer to the increase of BL humidity?

b) Investigate how does change of ET influence diurnal variation of the BL? How is the latter influence diurnal variations and intensity of deep convection? We will use the radiosound data and AMF1 and S-pol radar data to examine these questions. To separate the influences of land surface from AIE and large-scale circulation, we will compare cases with and without aerosols plumes and for different wind regimes.

c) Explore relative importance of ET, aerosols and monsoon circulation transition to increased convective intensity and transition from continental type to maritime type convection during the transition season: Regional model simulations with various land surface scenarios under different aerosols and large-scale dynamics lateral boundary conditions will be carried out and analyzed to explore this question in collaboration with other PIs.

Our results will also contribute to clarification of hypotheses I and IV.

26

Regional climatic impact of atmospheric teleconnections associated with Amazonian convection: sensitivity to model parameterizations and air-sea interaction PIs: R. Saravanan and Ping Chang, Texas A&M University Anticipated funding agencies: NSF-ATM Period: 1 January 2014, 3 years Estimated budget: $160K per year for modeling and data analysis Deficiencies in the simulation of Amazonian convection have been implicated as the cause of significant Atlantic climate biases in global coupled general circulation models. In particular, errors in the convective parameterizations used in models, as well as errors in boundary layer, land-surface and radiation parameterizations, can interact to produce overly strong or weak climatological precipitation in the Amazon region. The associated errors in diabatic heating can generate atmospheric teleconnections that lead to biases in surface wind stress over the tropical Atlantic Ocean. It has been hypothesized that the wind stress biases can be amplified by air-sea feedbacks and sub-surface oceanic processes, resulting in significant biases in the sea surface temperature (SST) simulations. The tropical Atlantic SST biases are closely tied to deficiencies in the rainfall simulation over semi-arid regions such as the northern Nordeste Brazil and the African Sahel. These model deficiencies significantly reduce our confidence in anthropogenic climate change projections in these vulnerable regions. The GoAmazon2014 field program affords a unique opportunity to improve the atmospheric model parameterizations that directly impact simulations of Amazon convection. The intense observations also provide an opportunity to design a rigorous test of the hypothesis that diabatic heating errors in this region lead to teleconnected atmospheric responses, amplified by oceanic feedbacks, that exert a significant influence on climate over the tropical Atlantic Ocean and adjoining continental regions. We propose to use a coupled regional climate model (CRCM), consisting of the WRF atmospheric model coupled to the ROMS ocean model. Using a regional model permits the use of high spatial resolution (up to 3km) compared to a typical global model at 50-100km resolution, thereby allowing explicit resolution of convective processes. It also excludes the remote influence of simulation errors in the Pacific region, which is often a complication encountered in global simulations. Seasonal ensembles of coupled and uncoupled simulations will be carried out over a region encompassing Amazonia, tropical Atlantic and portions of Africa. Different combinations of subgrid-scale parameterizations will be explored to identify the “control suite” of parameterizations that can best reproduce the statistical properties of field measurements of convective activity during GoAmazon2014. Sensitivity experiments will be carried out to assess the impact of deficiencies in Amazonian diabatic heating on rainfall patterns in Nordeste Brazil and western Africa, as well as the role of air-sea interaction in propagating these errors.

27

Observations and Modeling of the Green Ocean Amazon (GoAmazon2014) PI: Scot T. Martin, Harvard University Funding Agency: Department of Energy Main Deployment: 1 January 2014 through 31 December 2014 Budget: $3.6 million

The research activity seeks to understand how aerosol and cloud life cycles are influenced by pollutant outflow from a tropical megacity, particularly the susceptibility to cloud-aerosol-precipitation interactions and the feedbacks among biosphere and atmosphere functioning and human activities. The Mobile Aerosol Observing System (MAOS), the ARM Mobile Facility #1 (AMF1), several additional instruments that are not part of the standard package, and the G1 aircraft of the ARM Aerial Facility (AAF) are planned for deployment during 1 January 2014 through 31 December 2014. The main research site, nearby Manacapuru, Amazonas, Brazil, intersects the heavy pollution of the Manaus metropolis on a regular basis. The airshed of the site oscillates on cycles of two to three days between (i) one of the most pristine and natural continental sites on Earth and (ii) one heavily affected by tropical megacity pollution and its interactions with the forest natural emissions. For instance, particle number concentrations are at least 100 times greater in the pollution plume compared to the pristine conditions.

The scientific objectives of GoAmazon2014 respond to the question of, “What is the effect of pollution on natural atmosphere and ecosystem functioning and the couplings among them?” The experiment is organized around the carbon cycle, the aerosol life cycle, the cloud life cycle, and cloud-aerosol-precipitation Interactions, including but not limited to the following more detailed descriptions:

Carbon Cycle -improve Community Earth System Model (CESM) for land-atmosphere processes in the Amazon Basin, including aerosol-cloud-precipitation connections. Objective: Reduce uncertainties in our knowledge of feedbacks between vegetation-hydrology that underlie the Amazon forest dieback hypothesis. The uncertain range of feedbacks at present leads to large differences in ESM predictions. Objective: Response of photosynthesis and transpiration, including BVOC emissions, to changes in the direct and diffuse components of incoming solar radiation, i.e., in the context of current and future scenarios of aerosols and clouds in the Amazon Basin.

Aerosol Life Cycle -accurate modeling of aerosol sources/sinks and aerosol optical, cloud condensation nuclei (CCN), and ice nuclei (IN) properties, as affected by pollution of pristine tropical environments. Objective: The interactions of the urban pollution plume with biogenic volatile organic compounds in the tropics, especially the impact on the production of secondary organic aerosol, the formation of new particles, and biogenic emissions of aerosols and their precursors. Objective: Influence of anthropogenic activities on aerosol microphysical, optical, CCN, and IN properties in the tropics.

Cloud Life Cycle - development of a knowledge base to improve tropical cloud parameterizations in general circulation models (GCMs). Objective: The transition from shallow to deep cumulus convection during the daily cycle of the Amazon Basin, with comparison and understanding to other environments. Objective: The role of landscape heterogeneity—the Manaus urban area as well as the 10-km-scale of river width—on the dynamics of convection and clouds (+carbon cycle). Objective: The evolution of convective intensity from severe storms in the dry season to moderate storms in the wet season.

Cloud-Aerosol-Precipitation Interactions - improvement of parameterizations of aerosol-cloud interactions in climate models. Objective: Aerosol effects on deep convective clouds, precipitation, and lightning under different aerosol and synoptic regimes, including the roles of aerosols in changing regional climate and atmospheric circulation. Objective: Data-driven improvement of parameterizations of aerosol-cloud interactions in the climate models.

The theme uniting these objectives is the development of a data-driven knowledge base for predicting how the present-day functioning of energy, carbon, and chemical flows in the Amazon Basin might change, both due to external forcing on the Basin from global climate change and internal forcing from past and projected demographic and economic changes in the Basin. The ultimate goal is to estimate future changes in direct and indirect radiative forcing, energy distributions, regional climate, ecosystem functioning, and feedbacks to global climate.