Engineering topics
Publications and source records attributed to Chang, Alfred T. C..
Remote sensing research of snow cover has been accomplished for nearly 40 years. The use of visible, near-infrared, active and passive-microwave remote sensing for the analysis of snow cover is reviewed with an emphasis on the work on the last decade.
There are several hemispheric-scale satellite-derived snow-cover maps available, but none has been fully validated. For the period October 23 - December 25, 2000, we compare snow maps of North America derived from the Moderate Resolution Imaging Spectroradiometer (MODIS) and the National Oceanic and Atmospheric Administration (NOAA) National Operational Hydrologic Remote Sensing Center (NOHRSC), which both rely on satellite data from the visible and near-infrared parts of the spectrum; we also compare MODIS and Defense Meteorological Satellite Program (DMSP) Special Sensor Microwave/Imager (SSM/I) passive-microwave snow maps. The maps derived from visible and near-infrared data are more accurate for mapping snow cover than are the passive-microwave-derived maps, however discrepancies exist as to the location and extent of the snow cover among those maps. The large (approx. 30 km) footprint of the SSM/I data and the difficulty in distinguishing wet and shallow snow from wet or snow-free ground, reveal differences up to 5.32 million sq km in the amount of snow mapped using MODIS versus SSM/I data. Algorithms that utilize both visible and passive-microwave data, which would take advantage of the all-weather mapping ability of the passive-microwave data, will be refined following the launch of the Advanced Microwave Scanning Radiometer (AMSR) in the fall of 2001.
The Boreal Ecosystem-Atmosphere Study (BOREAS) Remote Sensing Science (RSS)-8 team utilized Landsat Thematic Mapper (TM) images to perform mapping of snow extent over the Southern Study Area (SSA). This data set consists of two Landsat TM images that were used to determine the snow-covered pixels over the BOREAS SSA on 18 Jan 1993 and on 06 Feb 1994. The data are stored in binary image format files. The RSS-08 snow map data are available from the Earth Observing System Data and Information System (EOSDIS) Oak Ridge National Laboratory (ORNL) Distributed Active Archive Center (DAAC). The data files are available on a CD-ROM (see document number 20010000884).
The surface meteorological data collected at the Boreal Ecosystem-Atmosphere Study (BOREAS) tower and ancillary sites are being used as inputs to an energy balance model to monitor the amount of snow storage in the boreal forest region. The BOREAS Hydrology (HYD)-2 team used Snow Water Equivalent (SWE) derived from an energy balance model and in situ observed SWE to compare the SWE inferred from airborne and spaceborne microwave data, and to assess the accuracy of microwave retrieval algorithms. The major external measurements that are needed are snowpack temperature profiles, in situ snow areal extent, and SWE data. The data in this data set were collected during February 1994 and cover portions of the Southern Study Area (SSA), Northern Study Area (NSA), and the transect areas. The data are available from BORIS as comma-delimited tabular ASCII files. The SWE data are available from the Earth Observing System Data and Information System (EOSDIS) Oak Ridge National Laboratory (ORNL) Distributed Active Archive Center (DAAC). The data files are available on a CD-ROM (see document number 20010000884).
We evaluated the performance of the Tropical Rainfall Measuring Mission (TRMM) Microwave Imager (TMI) at-launch algorithm for monthly oceanic rain rate using two years (January 1998 - December 1999) of TMI data. The TMI at-launch algorithm is based on Wilheit et al.'s technique for estimating monthly oceanic rainfall that relies on histograms of multichannel microwave measurements. Comparisons with oceanic monthly rain rates derived from the Defense Meteorological Satellite Program (DMSP) F-13 and F-14 Special Sensor Microwave Imager (SSM/I) data show the average rain rates over the TRMM region (between 400S and 40N) are 3.0, 2.85 and 2.89 mm/day, respectively for F-13, F-14 and TMI. Based on the latest version of TB data (version 5), both rainrate and freezing height derived from TMI are similar to those from the F-13 and F-14 SSM/I data. However, regionally the differences are statistically significant at the 95% confidence. Three hourly monthly rainrates are also computed from 3-hourly TB histograms to examine the diurnal cycle of precipitation. Over most of the oceanic TRMM area, a distinct early morning rainfall peak is found. A harmonic analysis shows that the amplitude of the 12h harmonic is significant and comparable to that of the 24h harmonic.
A technique for estimating monthly oceanic rainfall rate using multi-channel microwave measurements has been developed. This technique is applied to the Special Sensor Microwave Imager (SSM/I) data since 1987 to infer monthly rainfall for the Global Precipitation Climatology Project (GPCP). A modified version of this algorithm is now being applied to the TRMM Microwave Imager (TMI) data. There are three prominent features of this algorithm. First, the knowledge of the form of the rainfall intensity probability density function used to augment the measurements. Second, utilizing a linear combination of the 19.35 and 22.235 GHz channels to deemphasize the effect of water vapor. Third, an objective technique has been developed to estimate the rain layer thickness from the 19.35 and 22.23 5 GHz brightness temperature histograms, TMI data with better spatial resolution and 24 hour sampling (vs. sun-synchronized sampling, which is limited to two narrow intervals of local solar time for DMSP satellites) prompt us to study the similarity and difference between these two rainfall estimates. One year of rainfall data (January to December 1998) are used in this study. The mean differences of these two rain estimates are typically less than 15%. Statistical tests are administrated to evaluate the differences between rainfall estimates from SSM/I and TMI data. Differences between TMI and SSM/I rain estimates are discussed in the context of radiative transfer calculations and rainfall estimate algorithm.
In anticipation of the launch of the Earth Observing System (EOS) Terra, and the PM-1 spacecraft in 1999 and 2000, respectively, efforts are ongoing to determine errors of satellite-derived snow-cover maps. EOS Moderate Resolution Imaging Spectroradiometer (MODIS) and Advanced Microwave Scanning Radiometer-E (AMSR-E) snow-cover products will be produced. For this study we compare snow maps covering the same study area acquired from different sensors using different snow- mapping algorithms. Four locations are studied: 1) southern Saskatchewan; 2) a part of New England (New Hampshire, Vermont and Massachusetts) and eastern New York; 3) central Idaho and western Montana; and 4) parts of North and South Dakota. Snow maps were produced using a prototype MODIS snow-mapping algorithm used on Landsat Thematic Mapper (TM) scenes of each study area at 30-m and when the TM data were degraded to 1 -km resolution. National Operational Hydrologic Remote Sensing Center (NOHRSC) 1 -km resolution snow maps were also used, as were snow maps derived from 1/2 deg. x 1/2 deg. resolution Special Sensor Microwave Imager (SSM/1) data. A land-cover map derived from the International Geosphere-Biosphere Program (IGBP) land-cover map of North America was also registered to the scenes. The TM, NOHRSC and SSM/I snow maps, and land-cover maps were compared digitally. In most cases, TM-derived maps show less snow cover than the NOHRSC and SSM/I maps because areas of incomplete snow cover in forests (e.g., tree canopies, branches and trunks) are seen in the TM data, but not in the coarser-resolution maps. The snow maps generally agree with respect to the spatial variability of the snow cover. The 30-m resolution TM data provide the most accurate snow maps, and are thus used as the baseline for comparison with the other maps. Comparisons show that the percent change in amount of snow cover relative to the 3 0-m resolution TM maps is lowest using the TM I -km resolution maps, ranging from 0 to 40%. The highest percent change (less than 100%) is found in the New England study area, probably due to the presence of patchy snow cover. A scene with patchy snow cover is more difficult to map accurately than is a scene with a well-defined snowline such as is found on the North and South Dakota scene where the percent change ranged from 0 to 40%. There are also some important differences in the amount of snow mapped using the two different SSM/I algorithms because they utilize different channels.
Seventeen months of rainfall data (August 1987-December 1988) from nine satellite rainfall algorithms (Adler, Chang, Kummerow, Prabhakara, Huffman, Spencer, Susskind, and Wu) were analyzed to examine the uncertainty of satellite-derived rainfall estimates. The variability among algorithms, measured as the standard deviation computed from the ensemble of algorithms, shows regions of high algorithm variability tend to coincide with regions of high rain rates. Histograms of pattern correlation (PC) between algorithms suggest a bimodal distribution, with separation at a PC-value of about 0.85. Applying this threshold as a criteria for similarity, our analyses show that algorithms using the same sensor or satellite input tend to be similar, suggesting the dominance of sampling errors in these satellite estimates.
We examined nine satellite rainfall algorithms and compared the rain fields produced from these algorithms for the period of August 1987 to December 1988. Preliminary results show algorithms which use the same satellite sensor data tend to be similar, suggesting the importance of sampling. Oceanic global mean rainfall ranges from 2.7 to 3.6 mm/d. The variability in zonal mean rain rate is about 1.5-2 mm/d for these algorithms.
Global averages and random errors associated with the monthly oceanic rain rates derived from the Special Sensor Microwave/Imager (SSM/I) data using the technique developed by Wilheit et al. (1991) are computed. Accounting for the beam-filling bias, a global annual average rain rate of 1.26 m is computed. The error estimation scheme is based on the existence of independent (morning and afternoon) estimates of the monthly mean. Calculations show overall random errors of about 50-60 percent for each 5 deg x 5 deg box. The results are insensitive to different sampling strategy (odd and even days of the month). Comparison of the SSM/I estimates with raingage data collected at the Pacific atoll stations showed a low bias of about 8 percent, a correlation of 0.7, and an rms difference of 55 percent.
Three years of monthly rain rates over 5 deg x 5 deg latitude-longitude boxes have been calculated for oceanic regions 50 deg N-50 deg S from measurements taken by the Special Sensor Microwave/Imager on board the Defense Meteorological Satellite Program satellites using the technique developed by Wilheit et al. (1987, 1991). The annual and seasonal zonal-mean rain rates are larger than Jaeger's (1983) climatological estimates but are smaller than those estimated from the GOES precipitation index (GPI) for the same period. Regional comparison with the GPI showed that these rain rates are smaller in the north Indian Ocean and in the southern extratropics where the GPI is known to overestimate. The differences are also dominated by a jump at 170 deg W in the GPI rain rates across the mid-Pacific Ocean. This jump is attributed to the fusion of different satellite measurements in producing the GPI.
Multiple-channel microwave radiometric measurements made over Alaska at aircraft (near 90 and 183 GHz) and satellite (at 37 and 85 GHz) altitudes are used to study the effect of atmospheric absorption on the estimation of snow depth. The estimation is based on the radiative transfer calculations using an early theoretical model of Mie scattering of single-size particles. It is shown that the radiometric correction for the effect of atmospheric absorption is important even at 37 GHz for a reliable estimation of snow depth. Under a dry atmosphere and based on single-frequency radiometric measurements, the underestimation of snow depth could amount to 50 percent at 85 GHz and 20-30 percent at 37 GHz if the effect of atmospheric absorption is not taken into account. The snow depths estimated from the 90-GHz aircraft and 85-GHz satellite measurements are found to be in reasonable agreement. However, there is a discrepancy in the snow depths estimated from the 37-GHz (at both vertical and horizontal polarizations) and 85-GHz satellite measurements.
A study of differences between the morning and evening monthly rainfall for 5 x 5 deg cells over the oceans from the SSM/I data has been conducted. The monthly rainfalls are estimated from the technique given by Wilheit et al. The difference between the morning and evening monthly rainfall arises due to the various random errors involved in the retrieval process, the sampling error in the observations, and the diurnal component of oceanic rainfall. The diurnal component is weak but clearly visible when averaged over large areas and for long time periods. The analysis shows that morning rainfall is consistently greater than evening rainfall. The Northern Hemisphere seem to have a larger diurnal variation than does the Southern Hemisphere. The maximum ratio between the morning and evening monthly rainfall is 1.7 while 1.2 is the more typical value.
An algorithm for the estimation of monthly rain totals for 5 deg cells over the ocean from histograms of SSM/I brightness temperatures has been developed. There are three novel features to this algorithm. First, it uses knowledge of the form of the rainfall intensity probability density function to augment the measurements. Second, a linear combination of the 19.35 and 22.235 GHz channels has been employed to reduce the impact of variability of water vapor. Third, an objective technique has been developed to estimate the rain layer thickness from the 19.35- and 22.235-GHz brightness temperature histograms. Comparison with climatologies and the GATE radar observations suggest that the estimates are reasonable in spite of not having a beam-filling correction. By-products of the retrievals indicate that the SSM/I instrument noise level and calibration stability are quite good.
Explore the source record for details and available documents.
There will be six facility instruments on the NASA NPOP-1 and NPOP-2 and additional instruments on the Japanese and European satellites. Also, there are the 24 selected NASA instruments that may be flown on one of the platforms. Many of these instruments can provide data that could be very useful for real-time data studies in the snow and ice area. Any one instrument is not addressed in particular, but emphasis is placed on what is potentially possible using the capabilities of some of these instruments.
Satellite microwave data were used to evaluate the average areal water equivalent of snow cover in the mountainous Rio Grande basin of Colorado. Areal water equivalent data for the basin were obtained from contoured values of point measurements and from zonal water volume values generated by a snowmelt runoff model. Comparison of these snow water equivalent values shows the model values to consistently exceed the contoured values, probably because of the narrow elevation range in the lower part of the basin where the point measurements are concentrated. A significant relationship between the difference in microwave brightness temperatures at two different wavelengths and a basin-wide average snow water equivalent value is obtained. The average water equivalent of the snow cover in the basin was derived from differences of the microwave brightness temperatures.
Landsat-5 digital numbers have been used to compute the at-satellite planetary reflectance on spectrally similar zones on the Meares and Schwanda glaciers in Alaska and the Grossglockner glacier group in Austria. The patterns of TM-derived reflectances in the ablation areas of the Grossglockner glacier group and the Meares Glacier compare favorably with published reflectance curves measured on the surface of glacier ice, though the surface reflectance of snow-covered ice is higher than the Landsat-derived reflectance for the glaciers studied. The accuracy of the at-satellite planetary reflectances is shown to be affected by topographic and atmospheric effects and by the anisotropic nature of snow reflectance.