Gabriel Senay 2016-02-26 23:58:36
Global monitoring of weather, water, and crops with satellites and data integration Imagine a family of six whose livelihood is based on subsistence farming on a small, maybe one hectare, parcel of land somewhere in Africa. The seasonal rainfall varies greatly, from 500 to 800 mm per year, and the land is degraded. Thus, the parcel’s total productivity is not more than 1.5 tonnes in a good year, hardly meeting the family’s food requirements. The lack of surplus grain eliminates the need for grain storage structures, and due to the high rainfall variability, drought hazard is always looming, with an average recurrence interval of five to ten years. When severe drought occurs, famine can be averted only by external intervention, through a combination of national and international government support. In addition, because of differences in complex climatic factors, the impact of drought can be highly localized; thus, tools that can accurately monitor the severity and extent of drought are critical. Although the drought risk may be just one in ten years for a given location, drought will almost certainly occur somewhere in a given year. Continuous satellite-based monitoring of weather, crop performance, and water resources provides globally consistent data that are relevant for local decision-making, with enough lead time to mitigate the impact of drought hazards on vulnerable populations. In recent years, the usefulness of satellite data has increased for multiple applications—from the high-resolution mapping available through Google Earth to providing input data for hydrologic model parameterization and simulation. This article summarizes the integration of remotely sensed data and weather datasets for hydrologic modeling and its application for seasonal monitoring and early warning of food security risks, with a particular emphasis on the poor countries of the world. Global drought monitoring products are now regularly auto-generated for the U.S. Agency for International Development (USAID) Famine Early Warning System Network (FEWS NET) and posted online (http://earlywarning.usgs.gov/fews/product/109) in near-real time, as soon as satellite-acquired data are available. The spatially explicit (every 1 km2) data are made available in two forms: color-coded graphics that show drought stress levels, and image data that can be downloaded for further analysis by local experts. The convergence of evidence from three or more independently generated datasets can guide decision-makers to ensure that food aid is acquired and delivered in time to the right people. Satellite-derived data for rainfall, vegetation condition, and evapotranspiration (ET) are used regularly. While satellite-derived rainfall and vegetation condition indicators have been in use for more than ten years, the use of satellite-derived ET data for global monitoring is new because generating ET data required a simplification of the surface energy balance equations to handle the complexity of solving the energy balance terms for every pixel location around the world. This is an innovative aspect of the ET algorithm developed at the U.S. Geological Survey Earth Resources Observation and Science Center (USGS EROS). The basics of remote sensing Remote sensing is the science of collecting information without direct contact, such as with a camera. Although the human eye is, strictly speaking, a remote sensing organ, remote sensing technology for earth observation includes collecting, archiving, processing, and distributing data in a variety of forms for a wide variety of applications. Remote sensing applications typically produce images using specific portions of the electromagnetic (EM) spectrum (from visible to microwave frequencies) depending on the objective and the design characteristics of the sensor. Earth observation sensors are typically installed on aircraft or satellite platforms and acquire images with spatial resolutions ranging from <1 m to 1000 m per pixel and with a revisit cycle ranging from hourly to bi-weekly. While some sensors are considered active, generating their own EM signal and measuring the time delay of the returning signal, most sensors are passive and record the solar radiation reflected from the earth or other earth-emitted thermal or microwave signals. Satellites can also be classified as either geostationary (viewing the same location throughout the day as they move with the same angular speed as the earth) or polar orbiting (orbiting from pole to pole and revisiting the same location every few days depending on the desired spatial resolution). Interpretation of remote sensing images requires an understanding of the spectral signatures of the observed objects. For example, healthy vegetation looks green because it reflects relatively more of the green band of sunlight compared to its preferential absorption of the blue and red bands. Vegetation also reflects strongly in the infrared region of the EM spectrum. The levels of reflectance in different parts of the EM spectrum can be quantified mathematically and manipulated using image processing software to detect patterns of healthy and unhealthy vegetation. Similarly, the cold cloud duration (CCD) below a certain threshold temperature (about -60°C) can be used to estimate precipitation rates using geostationary satellites. ET modeling and mapping Recently, researchers have been developing ET models based on the surface energy balance to quantify the amount of water used by the landscape based on the amount of heat emitted by the earth. One such model developed at USGS EROS for national and international applications is the Operational Simplified Surface Energy Balance (SSEBop) model. As part of the energy transfer among the sun, the earth, and space, the earth emits stored energy to maintain a temperature that is based on a balanced energy budget between the incoming (sun to earth) and outgoing (earth to space) thermal energy. The earth-emitted thermal energy is a function of the surface temperature (Ts) and the surface emissivity, where emissivity is a property of materials that ranges in value from 0 to 1. Since satellite sensors can detect and quantify the heat emitted from the earth, with knowledge of the emissivity of the landscape, the surface temperature of the landscape can be determined. The Ts value is closely related to root zone soil moisture and ET, where ET is the transfer of water from the root system to the atmosphere through a combination of surface evaporation and transpiration. It is important to note the distinction between the surface temperature that the satellite estimates and the air temperature (Ta) that is measured about 1.5 m above the ground by local weather stations. Since ET involves a change of state in which liquid water is converted into water vapor, the process is accompanied by a large removal of energy from the landscape as heat is carried away with the evaporating water. Thus, at a given location, a relatively cooler land surface indicates a higher level of ET as compared to a warmer surface. For example, a dry, bare soil surface can be as much as 22°C warmer than a wet, well-vegetated surface during the peak crop growing season. Although solving the full energy balance of an evaporating surface is complex, the Ts value can be used to create an empirical index (0 to 1) by which the actual ET can be estimated as a fraction of the potential ET, which can be estimated from ambient weather parameters such as solar radiation, air temperature, relative humidity, wind, and air pressure. One of the challenges in using Ts as a measure of ET is that Ts varies across space not only due to water stress but also due to other factors, such as elevation changes. Thus, the method of establishing hot (dry) and cold (wet) reference limits through visual inspection only applies to limited areas. For a global extent, a new parameterization was developed to combine the satellite-derived surface temperature with the local air temperature. The two temperatures (Ts and Ta) tend to be close in magnitude when the pixel location is wellwatered vegetation, i.e., (Ts – Ta = 0). On the other hand, Ts – Ta becomes large, as much as 22°C, when the surface is bare and dry (fig. 1). With the introduction of air temperature into the model parameterization, SSEBop is currently being applied for monitoring every 1 km2 of the globe for ET anomalies, and updated every ten days (fig. 2). Thus, satellite-based sensors, such as Landsat and MODIS, provide the land surface temperate needed for estimating actual ET as a fraction of potential ET, and modelassimilated global weather datasets, such as the Global Data Assimilation System (GDAS), provide the weather parameters needed to determine potential ET. The integration of satellite data and weather datasets allows the determination of actual ET for the globe as soon as image data are acquired. For example, the latest ET anomaly map of the Horn of Africa (fig. 3) shows large areas with poor performance, i.e., seasonal ET more than 50% lower than the median as a result of failed rains. This has been corroborated by country-level and United Nations reports, with an estimated 4.5 million people needing emergency food aid in Ethiopia alone. With 1 km2 spatial resolution, the information from remote sensing is regularly consulted in the management and delivery of food aid at a sub-district scale. Satellite-derived ET also provides important information for quantifying irrigation water use and for understanding water availability as part of a region’s water budget. For example, figure 4 shows field-scale seasonal ET maps of the lower Colorado River basin derived using 100 m thermal imagery from Landsat 8. One of the unique contributions of remote sensing is that it has enabled collection of environmental data around the world with a uniform level of detail and accuracy, regardless of the economic development level of individual countries—thus becoming the great equalizer! Satellite data become even more useful when integrated with knowledge of hydrologic processes to produce value-added and management-relevant parameters, such as ET at multiple spatio-temporal scales, and globally consistent datasets. However, while satellitederived information is consistent, it is only an estimate, and it requires high-quality field information to correct potential biases. Therefore, users are encouraged to develop a satellite-integrated field information (SIFI) system that combines the synoptic coverage of satellite data with quality ground-truthing and calibration using on-site data. With cloud computing and increasingly sophisticated sensors, remote sensing from a variety of platforms, including unmanned aerial vehicles, will become an integral part of daily decision-making in agriculture, natural resources, and the environment. The scope of our discipline is broad, and we will continue to play a major role in the integration of remote sensing with decision-making tools and procedures, from precision agriculture to irrigation management and drought monitoring, with the aim of increasing efficiency, productivity, and resiliency in the waterfood-energy nexus. ASABE member Gabriel Senay, P.E., Research Physical Scientist, U.S. Geological Survey (USGS) Earth Resource Observation and Science (EROS) center, North Central Climate Science Center, Colorado State University, Fort Collins, USA, firstname.lastname@example.org. Senay is currently involved in monitoring crop performance in Africa, Central America, Southeast Asia, and the U.S. and developing new algorithms for computing evapotranspiration from remotely sensed data. Read more about his work in “How to predict a famine before it strikes” in Smithsonian Magazine (May 2015), available at www.smithsonianmag.com/innovation/predict-faminebefore- strikes-180954945/#QdjYRri2YcAfvyiR.99.
Published by ASABE. View All Articles.
This page can be found at http://bt.e-ditionsbyfry.com/article/The+Power+of+Remote+Sensing/2411236/292220/article.html.