The Hudson River estuary extends 246 km from the Upper New York Bay to Troy N.Y. and drains an area of 34,680 square kilometers. Per area nitrogen loading to the Hudson River estuary is the highest of any major estuary in the United States and has likely always been, even prior to large-scale anthropogenic inputs; however, the system has long been considered tolerant of such high loadings due to rapid flushing of the estuary by high water discharge from the Hudson River. This has been changing over the past decade as changes in the climate have led to lower freshwater discharge into the estuary and increased water residence times capable of supporting large algal blooms.

  • Overview
  • Silica dynamics
  • Urban environment
  • Atmospheric deposition
  • Climate
  • Downloads

The Howarth/ Marino lab has studied the Hudson basin, sub-basins, and estuary extensively since 1987. In our previous work, we have:

  • assessed rates of metabolism in both the tidal fresh water and saline regions of the estuary;
  • developed innovative approaches for measurements in the saline reach;
  • calculated the organic carbon budget for the tidal freshwater Hudson;
  • modeled the effect of climate and land use change on the flux of organic carbon from the landscape to the estuary;
  • assessed the influence of climate variability and freshwater discharge on estuarine primary production;
  • and compilled historical trends in how nutrient and carbon budgets for the saline estuary have changed through time.

Our current research aims to improve estimates of nutrient fluxes, including that of nitrogen (N; particulary the atmospheric deposition term) and silica (Si), and improve scientific understanding of the processes controlling these fluxes, including climate variability and land-use changes.

This project addresses the sources of nitrogen (N) inputs to the Hudson River estuary, the potential changes in riverine N flux as a result of climate change and variability, and the estuary's sensitivity to eutrophication in the face of changing climate and N loading. Our previous work on Hudson River nitrogen budgets has focused mainly on the more rural sub-basins of the Mohawk and Upper Hudson in part because of the large uncertainty over the magnitude of atmospheric deposition in urban areas. We are using a newly developed emission-based model (CMAQ) to improve estimates of atmospheric deposition across the entire watershed, as well as seasonal sampler deployments in the NYC metropolitan area to measure gas concentrations and estimate urban deposition. A combination of the NANI approach and ReNuMa modeling

While excess nitrogen (N) and phosphorus (P) have typically been held responsible for eutrophication problems, a report of the National Academy of Science’s Committee on Causes and Management of Coastal Eutrophication in 2000 noted the importance of silica (Si) delivery to estuaries, and the implications of Si limitation in the frequency of harmful algal blooms (HABs).

Diatoms, which require Si to form their characteristic exoskeletons, are considered relatively benign members of the algal community. They dominate many freshwater and estuarine ecosystems and serve as a critical food source for estuarine zooplankton, which in turn support larval fish production. When Si concentrations are so low that diatom growth is limited, diatoms may be outcompeted by non-siliceous algal species, including some species which exhibit undesirable and even toxic characteristics (Rabalais et al. 1996; NRC 1993). Evidence suggests that when Si:DIN molar ratios fall below 1, the food chain support declines (Turner et al. 1998). Where long-term data exist, decreases in Si relative to N or P have been correlated with increased occurrences of HABs (Ragueneau et al. 2006; Smayda 1990). Tracking the ratio of fluxes of N or P to Si provides a basis for assessing the relative risk of such occurrences.

Ultimately, the major source of dissolved Si (DSi) in rivers is rock weathering, though in densely populated regions, wastewater can also be a large source. Large reservoirs of biogenic Si also exist in soils, in grasses and other plant species, and to the extent that plants influence soil and rock weathering, they may accelerate Si cycling (Conley et al. 2002; Derry et al., 2005). Sinks include biological uptake and sedimentation in lakes, wetlands, or behind impoundments. It has previously been shown that biological uptake in combination with impoundments can preferentially deplete dissolved silica (Humborg et al 1997).The Hudson watershed is dominated by non-siliceous bedrock and is characterized by a long history of dam building; not surprisingly, Si concentrations are relatively low in the Hudson River (~20-45% of global average river concentrations; Simpson et al 2006), which may make it especially prone to toxic blooms.

Click on image to open full at full size

Most seasonal variation in the Hudson can be attributed to instream uptake by diatoms. Based on long term (span of ~50 years) USGS data for the Upper Hudson and Mohawk, seasonal variation of median monthly streamflows are fairly similar, but the strong seasonal variations of Si concentration differ qualitatively from the seasonal pattern of streamflow and from each other.  Similar winter maximum concentrations of SiO2 occur in both watersheds, suggesting that the net Si production in months of low biological activity is similar in both watersheds. Summertime concentrations differ substantially in both degree and timing, suggesting significant watershed-scale differences in instream retention. Previous estimates of watershed nutrient flux assume seasonally invariant nutrient retention; by characterizing the seasonal retention of Si and associated nutrients across watersheds, we are improving flux estimates. (USGS data:

By using existing information on Si to modify the Regional Nutrient Management (ReNuMa) model (previously developed in the Howarth/ Marino lab) we can better estimate nutrient flux to the Hudson estuarine system and potential impacts such as HAB occurrences.

In the Hudson, Si riverine flux can be treated as the result of the balance between bedrock sources and instream biotic uptake, though point sources within subwatersheds as well as retention by impoundments and seasonal variation of instream diatom productivity are accounted for. From a modeling standpoint, we plan to use spatial distributions of population density to estimate the wastewater contribution of DSi and existing rock weathering estimates to estimate type concentrations of Si from Si-bearing mineral deposits in the watershed.


Conley DJ. 2002. Terrestrial ecosystems and the global biogeochemical silica cycle. Global Biogeochem Cycles 16: 1121.

Derry LA, Kurtz AC, Ziegler K, and Chadwick OA. 2005. Biological control of terrestrial silica cycling and export fluxes to watersheds. Nature 433: 728–31.

National Research Council. 2000. Clean Coastal Waters:  Understanding and Reducing the Effects of Nutrient Pollution.  National Academy Press, Washington, DC.

National Research Council. 1993. Managing wastewater in coastal urban areas. National Academy press, Washington, DC.

Rabalais, N.N., R.E. Turner,  D. Justic, Q. Dortch, W.J. Wiseman, and B.K. Sen Gupta. 1996. Nutrient changes in the Mississippi River and system responses on the adjacent continental shelf. Estuaries 19:386-407

Ragueneau, O., D.J. Conley, A. Leynaert, S. Ni Longphuirt, and C.P. Slomp. 2006. Role of diatoms in silicon cycling and coastal marine food webs. Pp 163-195 in: V. Ittekot, D. Unger, C. Humborg, and N. Tac An (eds). The silicon cycle. Human perturbations and impacts on aquatic systems. SCOPE 66 Island Press, Washington, DC. 275 pp.

Simpson, H.J., S.N. Chillrud, R.F. Bopp, E. Shuster and D.A. Chaky. 2006. Major ion geochemistry and drinking water supply issues in the Hudson river basin. Pages 79-96 in Levinton, J. S., and J. R. Waldman (editors), The Hudson River Estuary, Cambridge Univ. Press.

Schwarz, G.E., Hoos, A.B., Alexander, R.B., and Smith, R.A., 2006, SPARROW-MOD: user documentation for the SPARROW surface water-quality model: U.S. Geological Survey Techniques and Methods, book 6, section B, Surface water, chapter 3 (6–B3).

Smayda, T. 1990. Novel and nuisance phytoplankton blooms in the sea: evidence for a global epidemia. Pp 29-40 in: E. Graneli (ed). Toxic Marine Phytoplankton. Elsevier, New York.

Swaney, D.P., K.E. Limburg and K. Stainbrook. 2006. Some Historical Changes in the Patterns of Population and Land Use in the Hudson River Watershed. In: J. Waldman, K.E. Limburg and D. Strayer, (eds.), Hudson River Fishes and their Environment. American Fisheries Society, Bethesda, MD. 365 pp.

Turner, R. E., N. Qureshi, N. N. Rabalais, Q. Dortch, D. Justic´, R. F. Shaw, and J. Cope. 1998. Fluctuating silicate:nitrate ratios and coastal plankton food webs. Proc. Ntl. Acad. Sci. 95:13048–13051.

USEPA. 2001. Better Assessment Science Integrating Point and Nonpoint Sources. BASINS Version 3.0. Users’ Manual. Office of Water. United States Environmental Protection Agency Report EPA 823B01001. Washington

van Bennekom, A.J. and W. Salomons. 1981. Pathways of nutrients and organic matter from land to ocean through rivers. In: River inputs to Ocean systems., ed J.M. Martin, J.D. Burton, and D. Eisma. Pp33-15. UNEP In. Ozone Comm., Sci. Comm. on Oceanic Res. Paris.

Funding for this project is provided by USGS and NYDEC Hudson River Environmental Program through New York State Water Resources Institute (WRI)

Historical Changes in the Food and Water Supply Systems of the New York Metropolitan Area

In September 2009, a diverse group of social and environmental researchers met at the University Pierre and Marie Curie (UPMC) in Paris for a workshop on the historical imprints of western cities and the interplay of developmental influence between urban and rural regions. Workshop discussions focused on select European and North American case studies (e.g. Paris, London, New York City) and addressed the following questions:

  • How did the development of cities in Europe and North America interact with the surrounding rural territories? How did these interactions change over time?
  • Did the development of the surrounding rural areas keep pace with that of cities?
  • Was there something in the relationships between these cities and their respective hinterland already pre-figuring the presently observed diverse development pattern before the inversion of the demographic ratio between urban and rural areas?
  • Can we still define the hinterland of a city in a globalized, web-connected world? What are the reasons to try re-linking cities to their surrounding territories?
  • Can we distinguish common trends or specific bifurcations in the historical trajectories of the relationships between cities and hinterlands?

Emission-based estimates for Total Nitrogen Deposition to the Landscape

Gases such as NO, NO2, and NH3 are not currently measured in any national monitoring programs; however, our previous work indicates that these nitrogen inputs may be large in urban, suburban, and industrialized areas. In a direct comparison of emission-based deposition estimates and those derived from monitoring data for the U.S. northeastern region (Maine to Virginia), we found that deposition estimates from emission-based models such as GCTM are 80% higher than those derived from monitoring data (Howarth 2006). We believe that the emission-based estimates are more reliable and that, because the monitoring networks do not measure nitrogen gases, estimates based on monitoring data may underestimate nitrogen inputs. However, emission-based models such as GCTM estimate deposition over large spatial scales too coarse to be applied to a single watershed. A new computationally intensive model developed at the University of Carolina Chapel Hill, the Community Multiscale Air Quality (CMAQ) model, is better able to assess deposition at these smaller spatial scales.

A major “hotspot” of the entire watershed for all forms of N deposition occurs at its southern limit, over the population center of the New York City metropolitan area. Of particular interest is the ratio of dry to wet N deposition estimated by the model. Most nitrogen deposition monitoring stations emphasize the measurement of wet deposition, assuming that corresponding dry deposition is relatively minor. We find that this can be an erroneous assumption, especially in dense urban areas with strong local emission sources - mostly motor vehicles.(Source: Howarth and Marino 2008)

We use output from CMAQ to estimate total N deposition over the Hudson and its sub-watersheds on a 36 km grid-cell basis. Constituents of the total N deposition have also been estimated, including the wet and dry deposition fractions as well as reduced and oxidized nitrogen constituents. Oxidized species of nitrogen are particularly important because of the longer atmospheric residence times, and thus greater distances between source and deposition.

The model for most of the watershed shows that oxidized dry N deposition is two or more times greater than reduced dry N deposition. Exceptions occur in highly urban-dominated regions (NYC metro area) and in agricultural areas (such as the Mohawk), due to strong local sources of NH3 from either agriculture (manure and fertilizer volatilization) or internal combustion engine (vehicle) emissions. Similar regional patterns are seen in the total (wet + dry) deposition. We have also developed aggregated deposition estimates for sub-watersheds of the Hudson/Mohawk, in order to evaluate the relative importance of atmospheric deposition in comparison to other anthropogenic inputs and riverine N fluxes.

We are currently working with EPA scientists from Research Triangle Park, NC, to examine higher spatial and temporal resolution deposition estimates (monthly, 12 km resolution) using the CMAQ model, which should provide more accurate aggregation at the sub-watershed level. A higher resolution model will also allow a better comparison with field measurements of dry N gas concentrations and associated dry N deposition estimates in the NYC area that we are making as part of this project.

Climate, Nitrogen Fluxes, and Sensitivity of the Estuary to Nutrient Pollution

Climate variation over the past few decades has already had a large and demonstrable impact on water quality in the Hudson River estuary, including reduced freshwater discharge leading to increased algal bloom events. Future climate change is expected to further impact estuarine nutrient responses as well as alter the rate of nutrient fluxes from the landscape (Moore et al 1997; Howarth et al 2006b). We are using an extension of the NANI method and the ReNuMa model to evaluate climate change impacts on N fluxes. We will then relate these changes to possible changes in estuarine response using the relationship between N load and primary productivity put forth in Nixon et al 1996 for water residence times > 1 day. Primary productivity is regulated by flushing and is generally low at residence times < 1 (Howarth et al 2006a) The simplest approach for evaluating the potential consequences of climate change on nitrogen fluxes is to use NANI budgeting. The newest version of the model, by incorporating data on precipitation and riverine discharge (Q), predicts riverine N flux with 89% accuracy.

Howarth et al. (2006b) found that, for the major watersheds of the northeastern U.S., the mean annual riverine nitrogen flux for 1988-1993 is very well explained as a function of NANI and discharge (left side). An equally good relationship is obtained using NANI and annual mean precipitation (right side). We can use these relationships to estimate hoe climate change may alter the sustained export of nitrogen to the Hudson River estuary (Source: Howarth et al 2006b)

While the NANI approach can provide an estimate of how climate may alter nitrogen fluxes from a large watershed on average over a period of many years, it can not address the impacts of extreme events, alterations of seasonal cycles, and inter-annual variability. Most climate change models predict greater intensification of the hydrologic cycle with more extreme precipitation event as well as greater drought periods and greater variance in weather (Moore et al 1997; Felzer and Heard 1999; Boesch et al 2000; Scavia et al 2002). To evaluate how these changes may affect nitrogen fluxes (plus freshwater discharge and sediment and phosphorus fluxes), we are using the ReNuMa model. ReNuMa provides robust flux estimates at monthly or seasonal scales, thus we can explore the interactive effects of land use change and climate variability on discharge and nutrient and sediment loads. To test the consequences of future climate change, we are using IPCC scenarios for predicted daily meteorological patterns.

Click on the icons to download files


  • Fluxmaster SiO2 model results 1989 - 1996:
  • xcel
  • Fluxmaster Nitrogen model results (DIN and TN) 1971 - 2006:
  • xcel
  • Nitrogen deposition (2002, CMAQ)
  • Net Anthropogenic Nitrogen Inputs (NANI)
  • Hudson Counties Hudson Catchments



  • Streamflow, 1971 - 2007:
  • xcel


  • Hudson/Mohawk Climate data, 1986 - 1993:
  • xcel
  • Hudson/Mohawk IPCC downscaled climate scenarios (a1b,a2,b1), daily,1961-2000:
  • a1b-temp  a1b-precip  a2-temp  a2-precip  b1-temp  b1-precip

         xcel             xcel           xcel          xcel            xcel          xcel


  • Hudson/Mohawk IPCC downscaled climate scenarios (a1b,a2,b1), daily, 2046-2065:

  • a1b-temp  a1b-precip  a2-temp  a2-precip  b1-temp  b1-precip

         xcel             xcel           xcel          xcel            xcel          xcel


  • Hudson/Mohawk IPCC downscaled climate scenarios (a1b,a2,b1), daily, 2081-2099

  • a1b-temp  a1b-precip  a2-temp  a2-precip  b1-temp  b1-precip

         xcel             xcel           xcel          xcel            xcel          xcel



Additional Spatial Datasets

  • Boundaries, Hudson xcel
  • Elevation, Mohawk basin (10x10)
  • Elevation, Upper Hudson (10x10)
  • Elevation, Hudson (30x30; NHD)
  • Dams, Hudson
  • Hydrography, Hudson catchments (NHD)
  • Land use, Hudson (2001, NLCD)
    • Agricultural land (%), Hudson
  • Lithology, Hudson xcel
  • Soil Topographic Index, Upper Hudson xcel
  • Soil Topographic Index, Mohawk xcel
  • Topographic Index, Upper Hudson xcel
  • Topographic Index, Mohawk xcel

Other Data Repositories (off site)

Return to Howarth/Marino Lab page