Nov 26 2008

Remote Sensing Data Assimilation

We have developed a computational framework for data assimilation of Fraction of Photosynthetically Active Radiation absorbed by vegetation (FPAR) and Leaf Area Index (LAI) from the MODerate Resolution Imaging Spectroradiometer (MODIS) to constrain empirical temperature, light, moisture and structural vegetation parameters in a prognostic phenology model.

Model-simulated FPAR (red curve) and LAI (blue curve) with a-posteriori parameters after the data assimilation compared to MODIS-based states (grey and black)

 

It uses the Ensemble Kalman Filter (EnKF) after Evensen [2003] employing around 1000 ensemble members as a sequential data assimilation model to assimilate satellite-derived vegetation states into the phenology model. The data assimilation is very effective in constraining model parameters which then yields a better a-posteriori prediction of vegetation states.

The method is firstly suitable to derive gap-free climatologies minimizing the errors of incomplete satellite retrievals, and it can be applied to parameterize prognostic vegetation models for use in prediction of vegetation states beyond the period of satellite coverage.

All code and related datasets are available as open source to the public:

http://phenoanalysis.sourceforge.net

R. Stockli, T. Rutishauser, D. Dragoni, J. O’Keefe, P. E. Thornton, M. Jolly, L. Lu, and A. S. Denning. Remote sens- ing data assimilation for a prognostic phenology model. Journal of Geophysical Research-Biogeosciences, 113(G4), Nov. 2008. doi: 10.1029/2008JG000781.

 

 

 

 

 

 

 

Jan 01 2007

Blue Marble images on the IPCC 4th assessment report (2007)

The Blue Marble western and eastern hemispheres by Reto Stöckli (NASA Earth Observatory) are displayed to the left as printed on the front cover of the IPCC (Intergovernmental Panel on Climate Change) 4th assessment WGI report “The Physical Science Basis”.

 

These images integrate land, ocean, sea ice and clouds into a visual representation of the earth’s climate system. They are based on space-borne earth observation data from NASA’s MODIS (MODerate resolution Imaging Spectroradiometer) sensor aboard the TERRA and AQUA satellites.

The visualizations of both hemispheres were created using the Blue Marble Next Generation (BMNG, 2005). The BMNG resolves 500 m per pixel and is the successor of the Blue Marble (2002) project providing a set of layers covering the earth’s climate system. Both datasets are still the most detailed continuous and seamless true-color images of the Earth’s surface ever produced.

I hope that you also feel inspired by these images and find them useful. They exemplify the comprehensive global observing capabilities of today’s state-of-the-art satellite sensors. They also display the beauty and diversity of the Earth as a combined and interrelated system.

Visual perception of Earth system dynamics can foster interest to further explore the underlying science. Furthermore such visualizations can help to increase public understanding and awareness of causes and effects of changes in Earth’s climate system.

The images and underlying datasets are provided at no cost to the public and there are no restrictions for commercial purposes. You’re welcome to send complimentary copies of products where you have applied these images. The high resolution  versions can be found here.

Citation: “Blue Marble Imagery by Reto Stöckli (NASA Earth Observatory) http://earthobservatory.nasa.gov “

Nov 27 2006

The European Heat wave of 2003 as seen from MODIS

Europe was experiencing a historic heat wave during the summer 2003. Compared to the long term climatological mean, temperatures in July 2003 were sizzling. Figure 1 below shows the differences in day time land surface temperatures of 2003 to the ones collected in 2000, 2001, 2002 and 2004 by the Moderate Resolution Imaging Spectroradiometer (MODIS) on NASA’s Terra satellite.

Figure: The visualization displays TERRA MODIS (MODerate resolution Imaging Spectroradiometer) derived land surface temperature data of 1km spatial resolution (Click on the image to get the high resolution TIFF file). The difference in land surface temperature is calculated by subtracting the average of all cloud free data during 2000, 2001, 2002 and 2004 from the ones in measured in 2003, covering the date range of July 20 – August 20. (Cite this image as: Image by Reto Stöckli, Robert Simmon and David Herring, NASA Earth Observatory, based on data from the MODIS land team).

Image story

A blanket of deep red across southern and eastern France (left of image center) shows where radiative temperatures were 10 degrees Celsius (18 degrees Fahrenheit) hotter in summer 2003 than in the other years. White areas show where temperatures were similar, and blue shows where temperatures were cooler in 2003 than in 2000,2001,2002 and 2004.

Even the Alps, which arc across southeastern France, Switzerland, Austria, and northern Italy (just below image center), were very warm this year. Glaciers were melting rapidly and swelling rivers and lakes to dangerously high levels. Climbers had to be evacuated from Switzerland’s famous Matterhorn after melting triggered the collapse of a rock face. The popular climbing destination had been closed while geologists assess the possibility of further collapses.

The heat wave stretched northward all the way to the United Kingdom, particularly southern England (bottom of island) and Scotland (top of island). In London, trains were shut down over fears that tracks would buckle in the heat, while in Scotland the high temperatures combined with falling water levels in rivers and streams are threatening the spawning and survival of salmon. Throughout France, Spain, Portugal, and Italy, the intense heat and dry conditions sparked devastating forest fires.

Download

Please find the highest resolution version of the 2003 LST, LAI and FPAR anomaly here.

References

Schär, C., P.L. Vidale, D. Lüthi, C. Frei, C. Häberli, M. Liniger and C. Appenzeller, 2004: The role of increasing temperature variability in European summer heat waves. Nature ,doi:10.1038/nature02300

Black, E., Blackburn, M., Harrison, G., Hoskins, B.J. and Methven, J. 2004. Factors contributing to the summer 2003 European heatwave. Weather, 59 , 217-223

R. Allen, Richard Lord Q.C., 2004: Climate change: the spectre of liability. Nature, 432, 551-552 (includes the above visualization).

Links

Jan 01 2006

Terra satellite launch & deploy (Remake from 2006)

For Terra’s 5th anniversary I have created a remake of the original launch & deploy 3D animation:

Dec 23 2005

Land use change in Sumatra

The Swiss NGO Paneco is engaged in Sumatra (Indonesia) where anthropogenic land use changes threaten local fauna and flora. The collaboration with Paneco has started with a small project that aimed at diagnosing rainfall intensity of a storm from TRMM 3B42 space radar data due to a lack of locally available rain gauges. The project continued with analysis of land use changes due to increasing oil palm plantations in northern Sumatra.

 

Mangroves woody trees or shrubs that grow in coastal areas and swamps. They occupy shallow water zones in tropical and subtropical coastal regions, usually where protected from direct wave action. The roots of the mangrove plants stabilize sand and mud. They also provide a habitat for wildlife, including several commercially important species of fish. Mangroves serve as a natural buffer to strong winds and waves produced by cyclones and they can protect against tsunamis.

 

image on the left was acquired by Landsat 5 in 1990 and shows a coastal mangrove area of around 400 km2 in south-western Aceh (Sumatra, Indonesia). The natural habitat is divided by the meandering Seumayam river which discharges into the Indian Ocean to the south. Dark red colors show natural vegetation, bright red indicate farming areas and blue colors mark bare soil and urban areas while blue-white colors are clouds. The image on the right was acquired by Landsat 7 in the year 2000 and shows the same area as above, but 10 years later: Almost all of the mangroves have been replaced with oil palm plantations. The rectangular and regular structures now show irrigation channels since oil palms have a high water demand. In the eastern part of the image a gradual expansion of farmland into the mangrove area can be witnessed.

Over 50% of the World’s mangroves have been lost during the past decades. In areas of the world where mangroves have been removed for development purposes, the coastline has been subject to rapid erosion. In Sumatra most of the mangrove trees have been replaced by oil palm plantations. Palm oil is a form of edible vegetable oil obtained from the fruit of the oil palm tree. Indonesia has become the second largest world producer of palm oil producing approximately 36% of world palm oil volume. There is increasing concern about environmental impacts of the palm oil industry since threatened species (such as the orang utan) live in the few remaining natural ecosystems on Borneo and Sumatra.

The false-color image from 5 August 2000 was acquired by the Enhanced Thematic Mapper plus (ETM+) aboard NASA’s Landsat 7 satellite. The one from 6 January 1990 was acquired by the Thematic Mapper (TM) aboard NASA’s Landsat 5 satellite. The Landsat satellites enable scientists to monitor land use and land cover change since 1972. Images by Reto Stöckli, based on Landsat data from the Global Land Cover Facility.

Further Reading

Paneco. Wanted: Undamaged Rainforest

Sumatran Orangutan Society

Earth Observatory: Deforestation in Indonesia

 

Jan 01 2005

Blue Marble Next Generation

The Blue Marble Next Generation (BMNG, 2005) is a cloud-free global satellite data set offering great spatial detail. It spans a longer data collection period than the original Blue Marble from 2000 and 2002. The original Blue Marble was a composite of four months of MODIS observations with a spatial resolution (level of detail) of 1 square kilometer per pixel. The Blue Marble Next Generation offers a year’s worth of monthly composites at a spatial resolution of 500 meters. These monthly images reveal seasonal changes to the land surface: the green-up and dying-back of vegetation in temperate regions such as North America and Europe, dry and wet seasons in the tropics, and advancing and retreating Northern Hemisphere snow cover.  Modified and shortened text from: http://bluemarble.nasa.gov

Space exploration changed our visual perception of planet Earth. In 1950s, satellites revolutionized weather forecasting when they began beaming home television images of cloud patterns. Astronaut photography in the early 1970s showed us the whole Earth“disk” for the first time in true-color – the so-called “Blue Marble”. Since 1972, satellite sensors have been acquiring atmosphere, land, ice and ocean data with increasing spectral, spatial and temporal resolution. Satellite remote sensing systems like NASAs Earth Observing System (EOS) help us to understand and monitor Earth’s physical, chemical, and biological processes.

One of our first false-color Earth images from back in the year 2000 is named the “Blue Marble”. That image was created for public outreach and it demonstrated the high value of such visualizations. New sensors like MODIS, aboard NASAs Terra and Aqua satellites, observe and measure a wide range of geophysical parameters. In 2002 we created a successor true-color Earth image using MODIS land, ocean, ice and atmosphere science products. This image has been widely used in museums, print media, TV documentaries, in movies, by mapping agencies, and in NASA’s public communications about its missions and research initiatives.

The success of the Blue Marble imagery motivated us to continue the project. The Blue Marble Next Generation (BMNG) is a cloud-free true-color dataset at 500-m spatial resolution and monthly temporal resolution. The BMNG aims at providing freely available data from the Earth surface in true color, derived from scientific data as a value-added product. Although the spatial resolution of the BMNG is comparable to other datasets, seasonal variations with monthly time-steps have not been shown before in seamless true-color composites. A visualization of seasonal variations (snowfall, droughts, wet seasons, spring greening, etc.) has good potential to enhance education. Furthermore, the BMNG can help to increase public understanding (and therefore acceptance) of satellite missions and awareness of causes and effects of changes in Earth’s climate system.

Visit the NASA’s official Blue Marble Next Generation Website: http://bluemarble.nasa.gov

Jan 01 2005

Blue Marble Next Generation Monthly Composites

A monthly collection of earth close-up images

Click on the respectiv images to get larger versions.

January 2004 Kenya

February 2004 Lake Etosha

March 2004 Sahara Desert

April 2004 The Alps

May 2004 The Himalayas

June 2004 The Andes

July 2004 Salt Lake City

August 2004 Eastern Greenland

September 2004 Patagonia

October 2004 The Okawango Delta

November 2004 Indonesian Islands

December 2004 The Bahamas

Oct 01 2004

PhD Thesis

Modeling and observations of seasonal land-surface heat and water exchanges at local and catchment scales over Europe (Doctoral thesis ETH No. 15742)

The land surface plays an important role in the global climate system, because it interacts dynamically with the atmosphere through manifold feedback mechanisms on a wide range of spatial and temporal scales. While on the one hand, weather and climate are known to influence vegetation phenology and its geographical distribution, soil and vegetation actively control land surface heat, water, momentum and carbon exchanges, thus influencing boundary layer development and convection. Evapotranspiration and runoff, in particular, which are balanced by precipitation, constitute the land portion of the water cycle, which is known to be a main contributor to climate variability. Knowledge about these processes and the ability to realistically model them is therefore of central importance in climate research. Simulated climate (and variability) are indeed sensitive to land surface parameterizations. There is, however, a gap between the local scale, at which land surface models and parameters are usually developed and evaluated, and the larger scales at which they are applied. This scale-gap needs to be bridged so that the high spatial and temporal dynamics of the land surface water cycle becomes part of modeled climate.

In order to help narrow the uncertainties in the modeling of seasonal-scale land-surface heat and water exchanges, local and catchment scale modeling experiments are performed in this study. Concurrently, different parameterizations are tested regarding their applicability in climate modeling, by exercising them on a wide range of climatic environments. All considered model formulations are embedded in a framework which includes ground and satellite remote sensing measurements, serving as an integration tool for the assessment of land surface processes. Satellite remote sensing is initially used to monitor vegetation state variables over Europe with a high temporal resolution, so that vegetation dynamics in land surface models can be prescribed with observed quantities. In a second stage local-scale measurements from FLUXNET are used for a process-based analysis of model results. Land surface models are applied at local scale using un-tuned large-scale and satellite-derived parameter sets. It is shown that soil storage processes play an important role in the seasonal heat and water fluxes and that vegetation biochemistry is a key component controlling the seasonal land surface water cycle. Finally, the Rhone-AGG initiative provides hydrological measurements on the catchment-scale, allowing for the exploration of scaling issues in the simulated water cycle. Catchment-scale simulations including lateral water fluxes, show that soil moisture drives runoff on the monthly time-scale and is largely controlled by evapotranspiration. While evapotranspiration was not found to be overly sensitive to runoff processes, the use of subgrid-scale topography-driven runoff provides a good simulation of the timing and magnitude of runoff at the daily to seasonal scale.

In summary, this study shows how satellite remote sensing, observations of boundary layer fluxes and ecosystem measurements can assist in developing models of the land surface water cycle which bridge the scale gaps between the processes involved; above-ground biophysics, relevant aspects of biochemistry and soil hydrology should be equally well represented in climate modeling applications.

Cover pictures: True color visualization of the cloud-free European land surface (August and February 2003): NASA’s Blue Marble Next Generation dataset, showing the seasonal changes in land surface reflectance, was derived from MODIS (MODerate resolution Imaging Spectroradiometer) MOD09A1 data by Reto Stöckli, NASA Earth Observatory.

The PDF of my PhD thesis is available here.

Jul 15 2004

Aura satellite launch & deploy

Aura is a earth observing satellite and it is part of NASA Earth Observing System (EOS). Aura’s four instruments study the atmosphere’s chemistry and dynamics. Aura’s measurements will enable us to investigate questions about ozone trends, air quality changes and their linkage to climate change.

Aura’s measurements will provide accurate data for predictive models and provide useful information for local and national agency decision support systems.

The Aura spacecraft was successfully launched on July 15, 2004 aboard a Delta II 7920-10L, a two stage expendable rocket, from the Vandenberg Western Test Range.

I have created the Aura launch and deploy animation for the Aura science team. The animations were used for public outreach purposes prior to launch.

 

Jan 01 2004

The European Fourier-adjusted and Interpolated NDVI (EFAI NDVI)

The EFAI NDVI is a high resolution land surface parameter dataset for climate modeling and mesoscale weather simulations. It was derived from the 20 year long NASA/NOAA AVHRR Pathfinder NDVI dataset, which is available in 10 day temporal and 0.1 degree spatial resolution for global area coverage.

The variability of vegetation state and function motivates the creation of high-resolution vegetation parameters varying dynamically over space and time. These parameters can be used to model the complex soil-vegetation-atmosphere interactions but also to assess the long-term changes in land use and vegetation physiology over a large area. The knowledge of the temporal variability of vegetation is a key requirement to quantify the carbon pools and exchanges on local and global scale. On global scale, satellite remote sensing now offers the possibility to estimate vegetation phenology with a high spatial and temporal revisiting frequency.

The NOAA/NASA Pathfinder dataset (James and Kalluri 1994) uses the NOAA polar orbiting satellites and is corrected for Rayleigh scattering and ozone absorption as well as for instrument degradation. It is however not corrected for aerosols or viewing geometry effects. High solar zenith angles in high latitudes causes long term data dropouts during winter. The cloud screening is problematic and leads to temporal inconsistencies in the original Pathfinder dataset. To create a spatio-temporally consistent vegetation phenology useful for climate research we used the Pathfinder NDVI (Normalized Difference Vegetation Index) over Europe and applied the following correction methodology:

  • Replacement of processing artifacts and data gaps in the dataset by spatial interpolation
  • Adjustment of the NDVI time-series by using a temporal interpolation procedure

The temporal interpolation procedure is a modification of Los (1998) and Sellers et al. (1996) and uses second order discrete fourier series to extract the seasonal variability of vegetation state and function from the Pathfinder NDVI dataset. The resulting spatio-temporally consistent 0.1 degree and 10 day NDVI dataset is called EFAI NDVI (European Fourier-Adjusted and Interpolated NDVI). A full description of this methodology is available in Stöckli and Vidale (2004). Biophysical land surface parameters were derived from the EFAI NDVI by applying simple empirical relationships between satellite radiometry and vegetation physiology.

The EFAI Dataset has also been processed at global scale and is available here. It has been superseded by higher quality and better gap-filled products. For instance our NASA Energy and Water Cycle MODIS-based FPAR/LAI data assimilation project called Pheno Analysis is a suitable successor of the EFAI NDVI.

Reference: R. Stöckli and P. L. Vidale (2004). European plant phenology and climate as seen in a 20-year avhrr land-surface parameter dataset, International Journal of Remote Sensing, 25(17), 3303–3330