Atmospheric radiative transfer models (RTMs) simulate the light propagation in the Earth's atmosphere. With the evolution of RTMs, their increase in complexity makes them impractical in routine processing such as atmospheric correction. To overcome their computational burden, standard practice is to interpolate a multidimensional lookup table (LUT) of prestored simulations. However, accurate interpolation relies on large LUTs, which still implies large computation times for their generation and interpolation. In recent years, emulation has been proposed as an alternative to LUT interpolation. Emulation approximates the RTM outputs by a statistical regression model trained with a low number of RTM runs. However, a concern is whether the emulator reaches sufficient accuracy for atmospheric correction. Therefore, we have performed a systematic assessment of key aspects that impact the precision of emulating MODTRAN: 1) regression algorithm; 2) training database size; 3) dimensionality reduction (DR) method and a number of components; and 4) spectral resolution. The Gaussian processes regression (GPR) was found the most accurate emulator. The principal component analysis remains a robust DR method and nearly 20 components reach sufficient precision. Based on a database of 1000 samples covering a broad range of atmospheric conditions, GPR emulators can reconstruct the simulated spectral data with relative errors below 1% for the 95th percentile. These emulators reduce the processing time from days to minutes, preserving sufficient accuracy for atmospheric correction and providing model uncertainties and derivatives. We provide a set of guidelines and tools to design and generate accurate emulators for satellite data processing applications.

Physically based radiative transfer models (RTMs) are widely used in Earth observation to understand the radiation processes occurring on the Earth's surface and their interactions with water, vegetation, and atmosphere. Through continuous improvements, RTMs have increased in accuracy and representativity of complex scenes at expenses of an increase in complexity and computation time, making them impractical in various remote sensing applications. To overcome this limitation, the common practice is to precompute large lookup tables (LUTs) for their later interpolation. To further reduce the RTM computation burden and the error in LUT interpolation, we have developed a method to automatically select the minimum and optimal set of input-output points (nodes) to be included in an LUT. We present the gradient-based automatic LUT generator algorithm (GALGA), which relies on the notion of an acquisition function that incorporates: 1) the Jacobian evaluation of an RTM and 2) the information about the multivariate distribution of the current nodes. We illustrate the capabilities of GALGA in the automatic construction and optimization of MODTRAN-based LUTs of different dimensions of the input variables space. Our results indicate that when compared with a pseudorandom homogeneous distribution of the LUT nodes, GALGA reduces:1) the LUT size by >24%; 2) the computation time by 27%; and 3) the maximum interpolation relative errors by at least 10%. It is concluded that an automatic LUT design might benefit from the methodology proposed in GALGA to reduce interpolation errors and computation time in computationally expensive RTMs.

Monitoring the flow of radiative energy at top-of-atmosphere (TOA) is essential for understanding the Earth's climate and how it is changing with time. The determination of TOA global net radiation budget using broadband nonscanner instruments has received renewed interest recently due to advances in both instrument technology and the availability of small satellite platforms. The use of such instruments for monitoring Earth's radiation budget was attempted in the past from satellite missions such as the Nimbus 7 and the Earth Radiation Budget Experiment (ERBE). This paper discusses the important lessons learned from the operation of the ERBE nonscanner instrument and the production of the ERBE nonscanner TOA radiation budget data set that have direct relevance to current nonscanner instrument efforts.

Previous research has revealed inconsistencies between the Collection 5 (C5) calibrations of certain channels common to the Terra and Aqua MODerate-resolution Imaging Spectroradiometers (MODIS). To achieve consistency between the Terra and Aqua MODIS radiances used in the Clouds and the Earth's Radiant Energy System (CERES) Edition 4 (Ed4) cloud property retrieval system, adjustments were developed and applied to the Terra C5 calibrations for channels 1-5, 7, 20, and 26. These calibration corrections were developed independently of those used for MODIS Collection 6 (C6) data, which became available after the CERES Ed4 processing had commenced. The comparisons demonstrate that the corrections applied to the Terra C5 data for CERES Edition 4 generally resulted in Terra-Aqua radiance consistency that is as good as or better than that of the C6 datasets. The C5 adjustments resulted in more consistent Aqua and Terra cloud property retrievals than seen in the previous CERES edition. Other calibration artifacts were found in one of the corrected channels and in some of the uncorrected thermal channels after Ed4 began. Where corrections were neither developed nor applied, some artifacts are likely to have been introduced into the Ed4 cloud property record. For example, the degradation in the Aqua MODIS 0.65-μm channel in both the C5 and C6 datasets affects trends in cloud optical depth retrievals. Thus, despite the much-improved consistency achieved for the Terra and Aqua datasets in Ed4, the CERES Ed4 cloud property datasets should be used cautiously for cloud trend studies because of those remaining calibration artifacts.

NASA's Ice, Cloud, and land Elevation Satellite (ICESat), which operated between 2003 and 2009, made the first satellite-based global lidar measurement of Earth's ice sheet elevations, sea-ice thickness and vegetation canopy structure. The primary instrument on ICESat was the Geoscience Laser Altimeter System (GLAS), which measured the distance from the spacecraft to Earth's surface via the roundtrip travel time of individual laser pulses. GLAS utilized pulsed lasers and a direct detection receiver consisting of a silicon avalanche photodiode (Si APD) and a waveform digitizer. Early in the mission, the peak power of the received signal from snow and ice surfaces was found to span a wider dynamic range than planned, often exceeding the linear dynamic range of the GLAS 1064-nm detector assembly. The resulting saturation of the receiver distorted the recorded signal and resulted in range biases as large as ~50 cm for ice and snow-covered surfaces. We developed a correction for this "saturation range bias" based on laboratory tests using a spare flight detector, and refined the correction by comparing GLAS elevation estimates to those derived from Global Positioning System (GPS) surveys over the calibration site at the salar de Uyuni, Bolivia. Applying the saturation correction largely eliminated the range bias due to receiver saturation for affected ICESat measurements over Uyuni and significantly reduced the discrepancies at orbit crossovers located on flat regions of the Antarctic ice sheet.

A conical cavity has been designed and fabricated for use as a broadband passive microwave calibration source, or blackbody, at the National Institute of Standards and Technology. The blackbody will be used as a national primary standard for brightness temperature and will allow for the prelaunch calibration of spaceborne radiometers and calibration of ground-based systems to provide traceability among radiometric data. The conical geometry provides performance independent of polarization, minimizing reflections, and standing waves, thus having a high microwave emissivity. The conical blackbody has advantages over typical pyramidal array geometries, including reduced temperature gradients and excellent broadband electromagnetic performance over more than a frequency decade. The blackbody is designed for use between 18 and 230 GHz, at temperatures between 80 and 350 K, and is vacuum compatible. To approximate theoretical blackbody behavior, the design maximizes emissivity and thus minimizes reflectivity. A newly developed microwave absorber is demonstrated that uses cryogenically compatible, thermally conductive two-part epoxy with magnetic carbonyl iron (CBI) powder loading. We measured the complex permittivity and permeability properties for different CBI-loading percentages; the conical absorber is then designed and optimized with geometric optics and finite-element modeling, and finally, the reflectivity of the resulting fabricated structure is measured. We demonstrated normal incidence reflectivity considerably below -40 dB at all relevant remote sensing frequencies.

A robust physics-based combined radar-radiometer, or Active-Passive, surface soil moisture and roughness estimation methodology is presented. Soil moisture and roughness retrieval is performed via optimization, i.e., minimization, of a joint objective function which constrains similar resolution radar and radiometer observations simultaneously. A data-driven and noise-dependent regularization term has also been developed to automatically regularize and balance corresponding radar and radiometer contributions to achieve optimal soil moisture retrievals. It is shown that in order to compensate for measurement and observation noise, as well as forward model inaccuracies, in combined radar-radiometer estimation surface roughness can be considered a free parameter. Extensive Monte-Carlo numerical simulations and assessment using field data have been performed to both evaluate the algorithm's performance and to demonstrate soil moisture estimation. Unbiased root mean squared errors (RMSE) range from 0.18 to 0.03 cm3/cm3 for two different land cover types of corn and soybean. In summary, in the context of soil moisture retrieval, the importance of consistent forward emission and scattering development is discussed and presented.

The NASA Soil Moisture Active Passive (SMAP) satellite mission was launched on January 31, 2015 to provide global mapping of high-resolution soil moisture and freeze-thaw state every 2-3 days using an L-band (active) radar and an L-band (passive) radiometer. The Level 2 radiometer-only soil moisture product (L2_SM_P) provides soil moisture estimates posted on a 36-km Earth-fixed grid using brightness temperature observations from descending passes. This paper provides the first comparison of the validated-release L2_SM_P product with soil moisture products provided by the Soil Moisture and Ocean Salinity (SMOS), Aquarius, Advanced Scatterometer (ASCAT), and Advanced Microwave Scanning Radiometer 2 (AMSR2) missions. This comparison was conducted as part of the SMAP calibration and validation efforts. SMAP and SMOS appear most similar among the five soil moisture products considered in this paper, overall exhibiting the smallest unbiased root-mean-square difference and highest correlation. Overall, SMOS tends to be slightly wetter than SMAP, excluding forests where some differences are observed. SMAP and Aquarius can only be compared for a little more than two months; they compare well, especially over low to moderately vegetated areas. SMAP and ASCAT show similar overall trends and spatial patterns with ASCAT providing wetter soil moistures than SMAP over moderate to dense vegetation. SMAP and AMSR2 largely disagree in their soil moisture trends and spatial patterns; AMSR2 exhibits an overall dry bias, while desert areas are observed to be wetter than SMAP.

The inter-comparison of MODIS reflective solar bands onboard Aqua and Terra is very important for assessment of each instrument's calibration. One of the limitations is the lack of simultaneous nadir overpasses. Their measurements over a selected Earth view target have significant differences in solar and view angles, which magnify the effects of atmospheric scattering and Bidirectional Reflectance Distribution Function (BRDF). In this work, an inter-comparison technique is formulated after correction for site's BRDF and atmospheric effects. The reflectance measurements over Libya desert sites 1, 2, and 4 from both the Aqua and Terra MODIS are regressed to a BRDF model with an adjustable coefficient accounting for calibration difference. The ratio between Aqua and Terra reflectance measurements are derived for bands 1 to 9 and the results from different sites show good agreement. For year 2003, the ratios are in the range of 0.985 to 1.010 for band 1 to 9. Band 3 shows the lowest ratio 0.985 and band 1shows the highest ratio 1.010. For the year 2014, the ratio ranges from approximately 0.983 for bands 2 and 1.012 for band 8. The BRDF corrected reflectance for the two instruments are also derived for every year from 2003 to 2014 for stability assessment. Bands 1 and 2 show greater than 1% differences between the two instruments. Aqua bands 1 and 2 show downward trends while Terra bands 1 and 2 show upward trends. Bands 8 and 9 of both Aqua and Terra show large variations of reflectance measurement over time.

The MODIS instruments aboard Terra and Aqua satellites are radiometrically calibrated on-orbit with a set of on-board calibrators (OBC) including a solar diffuser (SD), a blackbody (BB) and a space view (SV) port through which the detectors can view the dark space. As a whisk-broom scanning spectroradiometer, thirty-six MODIS spectral bands are assembled in the along-scan direction on four focal plane assemblies (FPA). These bands capture images of the same target sequentially with the motion of a scan mirror. Then the images are co-registered on board by delaying appropriate band-dependent amount of time depending on the band locations on the FPA. While this co-registration mechanism is functioning well for the "far field" remote targets such as Earth view (EV) scenes or the Moon, noticeable band-to-band misregistration in the along-scan direction has been observed for "near field" targets, in particular the OBCs. In this paper, the misregistration phenomenon is presented and analyzed. It is concluded that the root cause of the misregistration is that the rotating element of the instrument, the scan mirror, is displaced from the focus of the telescope primary mirror. The amount of the misregistration is proportional to the band location on the FPA and is inversely proportional to the distance between the target and the scan mirror. The impact of this misregistration to the calibration of MODIS bands is discussed. In particular, the calculation of the detector gain coefficient of bands 8-16 (412 nm - 870 nm) is improved by up to 1.5% for Aqua MODIS.

Repeat Image Feature Tracking (RIFT) is commonly used to measure glacier surface motion from pairs of images, most often utilizing normalized cross correlation (NCC). The Multiple-Image Multiple-Chip (MIMC) algorithm successfully employed redundant matching (i.e. repeating the matching process over each area using varying combinations of settings) to increase the matching success rate. Due to the large number of repeat calculations, however, the original MIMC algorithm was slow and still prone to failure in areas of high shearing flow. Here we present several major updates to the MIMC algorithm that increase both speed and matching success rate. First, we include additional redundant measurements by swapping the image order and matching direction; a process we term Quadramatching. Second, we utilize a priori ice velocity information to confine the NCC search space through a system we term dynamic linear constraint (DLC), which substantially reduces the computation time and increases the rate of successful matches. Additionally, we develop a novel post-processing algorithm, pseudosmoothing, to determine the most probable displacement. Our tests reveal the complimentary and multiplicative nature of these upgrades in their improvement in overall MIMC performance.

The Soil Moisture Active-Passive (SMAP) L-band microwave radiometer is a conical scanning instrument designed to measure soil moisture with 4% volumetric accuracy at 40-km spatial resolution. SMAP is NASA's first Earth Systematic Mission developed in response to its first Earth science decadal survey. Here, the design is reviewed and the results of its first year on orbit are presented. Unique features of the radiometer include a large 6-m rotating reflector, fully polarimetric radiometer receiver with internal calibration, and radio-frequency interference detection and filtering hardware. The radiometer electronics are thermally controlled to achieve good radiometric stability. Analyses of on-orbit results indicate that the electrical and thermal characteristics of the electronics and internal calibration sources are very stable and promote excellent gain stability. Radiometer NEDT 1 K for 17-ms samples. The gain spectrum exhibits low noise at frequencies 1 MHz and 1/f noise rising at longer time scales fully captured by the internal calibration scheme. Results from sky observations and global swath imagery of all four Stokes antenna temperatures indicate that the instrument is operating as expected.

Radiometers operating at L-band (1.4 GHz) are used to retrieve sea surface salinity over ice-free oceans and have been used recently to study the cryosphere. One hindrance of their use in the high latitudes is the preponderance of mixed scenes, where seawater and sea ice are both present in the sensor's field of view (FOV). Accurately characterizing the scene is crucial for oceanographic and cryospheric applications. To that end, a sea ice fraction model, composed of passive microwave sea ice concentration retrievals and an instrument simulator that integrates radiative power coming from all around the antenna, is used. We investigate the model currently used operationally to derive the ice fraction affecting the Aquarius observations and show that it can be significantly improved. On the one hand, the current model tends to overestimate sea ice fraction in the marginal ice zone where observations are used for salinity retrievals. On the other hand, the current model underestimates ice fraction within the ice pack where observations are used to derive sea ice properties. For the northern hemisphere, we also find evidence of the sea ice type impact on L-band radiometric observations. We present a model to derive sea ice fractions that are in better agreement with Aquarius radiometric observations using the Advanced Microwave Scanning Radiometer 2 Bootstrap algorithm for sea ice concentration and using high-resolution integration over the sensor's FOV.

The MODIS Level-2 cloud product (Earth Science Data Set names MOD06 and MYD06 for Terra and Aqua MODIS, respectively) provides pixel-level retrievals of cloud-top properties (day and night pressure, temperature, and height) and cloud optical properties (optical thickness, effective particle radius, and water path for both liquid water and ice cloud thermodynamic phases-daytime only). Collection 6 (C6) reprocessing of the product was completed in May 2014 and March 2015 for MODIS Aqua and Terra, respectively. Here we provide an overview of major C6 optical property algorithm changes relative to the previous Collection 5 (C5) product. Notable C6 optical and microphysical algorithm changes include: (i) new ice cloud optical property models and a more extensive cloud radiative transfer code lookup table (LUT) approach, (ii) improvement in the skill of the shortwave-derived cloud thermodynamic phase, (iii) separate cloud effective radius retrieval datasets for each spectral combination used in previous collections, (iv) separate retrievals for partly cloudy pixels and those associated with cloud edges, (v) failure metrics that provide diagnostic information for pixels having observations that fall outside the LUT solution space, and (vi) enhanced pixel-level retrieval uncertainty calculations. The C6 algorithm changes collectively can result in significant changes relative to C5, though the magnitude depends on the dataset and the pixel's retrieval location in the cloud parameter space. Example Level-2 granule and Level-3 gridded dataset differences between the two collections are shown. While the emphasis is on the suite of cloud optical property datasets, other MODIS cloud datasets are discussed when relevant.

This paper evaluates the retrieval of soil moisture in the top 5-cm layer at 3-km spatial resolution using L-band dual-copolarized Soil Moisture Active-Passive (SMAP) synthetic aperture radar (SAR) data that mapped the globe every three days from mid-April to early July, 2015. Surface soil moisture retrievals using radar observations have been challenging in the past due to complicating factors of surface roughness and vegetation scattering. Here, physically based forward models of radar scattering for individual vegetation types are inverted using a time-series approach to retrieve soil moisture while correcting for the effects of static roughness and dynamic vegetation. Compared with the past studies in homogeneous field scales, this paper performs a stringent test with the satellite data in the presence of terrain slope, subpixel heterogeneity, and vegetation growth. The retrieval process also addresses any deficiencies in the forward model by removing any time-averaged bias between model and observations and by adjusting the strength of vegetation contributions. The retrievals are assessed at 14 core validation sites representing a wide range of global soil and vegetation conditions over grass, pasture, shrub, woody savanna, corn, wheat, and soybean fields. The predictions of the forward models used agree with SMAP measurements to within 0.5 dB unbiased-root-mean-square error (ubRMSE) and -0.05 dB (bias) for both copolarizations. Soil moisture retrievals have an accuracy of 0.052 m/m ubRMSE, -0.015 m/m bias, and a correlation of 0.50, compared to measurements, thus meeting the accuracy target of 0.06 m/m ubRMSE. The successful retrieval demonstrates the feasibility of a physically based time series retrieval with L-band SAR data for characterizing soil moisture over diverse conditions of soil moisture, surface roughness, and vegetation.

The Soil Moisture and Ocean Salinity (SMOS) satellite provides retrievals of soil moisture in the upper 5 cm with a 30-50 km resolution and a mission accuracy requirement of 0.04 cm cm. These observations can be used to improve land surface model soil moisture states through data assimilation. In this paper, SMOS soil moisture retrievals are assimilated into the Noah land surface model via an Ensemble Kalman Filter within the NASA Land Information System. Bias correction is implemented using Cumulative Distribution Function (CDF) matching, with points aggregated by either land cover or soil type to reduce sampling error in generating the CDFs. An experiment was run for the warm season of 2011 to test SMOS data assimilation and to compare assimilation methods. Verification of soil moisture analyses in the 0-10 cm upper layer and root zone (0-1 m) was conducted using measurements from several observing networks in the central and southeastern United States. This experiment showed that SMOS data assimilation significantly increased the anomaly correlation of Noah soil moisture with station measurements from 0.45 to 0.57 in the 0-10 cm layer. Time series at specific stations demonstrate the ability of SMOS DA to increase the dynamic range of soil moisture in a manner consistent with station measurements. Among the bias correction methods, the correction based on soil type performed best at bias reduction but also reduced correlations. The vegetation-based correction did not produce any significant differences compared to using a simple uniform correction curve.

The MODerate-resolution Imaging Spectroradiometer (MODIS) is a premier Earth observing sensor of the early 21 Century, flying on-board the Terra (T) and Aqua (A) spacecrafts. Both instruments far exceeded their 6 year design life and continue to operate satisfactorily for more than 15 and 13 years, respectively. The MODIS instrument is designed to make observations at nearly a 100% duty cycle covering the entire Earth in less than 2 days. The MODIS sensor characteristics include a spectral coverage from 0.41 μm - 14.4 μm, of which those wavelengths ranging from 3.7 μm - 14. 4 μm cover the thermal infrared region which is interspaced in 16 Thermal Emissive Bands (TEB). Each of the TEB contains 10 detectors which record samples at a spatial resolution of 1 km. In order to ensure a high level of accuracy for the TEB measured Top Of Atmosphere (TOA) radiances, an onboard BlackBody (BB) is used as the calibration source. This paper reports the noise characterization and performance of the TEB on various counts. First, the stability of the onboard BB is evaluated to understand the effectiveness of the calibration source. Next, key noise metrics such as the Noise Equivalent Temperature difference (NEdT) and the Noise Equivalent dn difference (NEdN) for the various TEB are determined from multiple temperature sources. These sources include the nominally controlled BB temperature of 290 K for T-MODIS and 285 K for A-MODIS, as well as a BB Warm Up - Cool Down (WUCD) cycle that is performed over a temperature range from roughly 270 K - 315 K. The Space View (SV) port that measures the background signal serves as a viable cold temperature source for measuring noise. In addition, a well characterized Earth View (EV) Target, the Dome C site located in the Antarctic plateau, is used for characterizing the stability of the sensor, indirectly providing a measure of the NEdN. Based on this rigorous characterization, a list of the noisy and inoperable detectors for the TEB for both instruments is reported to provide the science user communities quality control of the MODIS Level 1B calibrated product.

This paper considers some of the issues of radiometer brightness image formation and reconstruction for use in the NASA-sponsored Calibrated Passive Microwave Daily Equal-Area Scalable Earth Grid 2.0 Brightness Temperature Earth System Data Record project, which generates a multisensor multidecadal time series of high-resolution radiometer products designed to support climate studies. Two primary reconstruction algorithms are considered: the Backus-Gilbert approach and the radiometer form of the scatterometer image reconstruction (SIR) algorithm. These are compared with the conventional drop-in-the-bucket (DIB) gridded image formation approach. Tradeoff study results for the various algorithm options are presented to select optimum values for the grid resolution, the number of SIR iterations, and the BG gamma parameter. We find that although both approaches are effective in improving the spatial resolution of the surface brightness temperature estimates compared to DIB, SIR requires significantly less computation. The sensitivity of the reconstruction to the accuracy of the measurement spatial response function (MRF) is explored. The partial reconstruction of the methods can tolerate errors in the description of the sensor measurement response function, which simplifies the processing of historic sensor data for which the MRF is not known as well as modern sensors. Simulation tradeoff results are confirmed using actual data.

Automatic image registration is the process of aligning two or more images of approximately the same scene with minimal human assistance. Wavelet-based automatic registration methods are standard, but sometimes are not robust to the choice of initial conditions. That is, if the images to be registered are too far apart relative to the initial guess of the algorithm, the registration algorithm does not converge or has poor accuracy, and is thus not robust. These problems occur because wavelet techniques primarily identify isotropic textural features and are less effective at identifying linear and curvilinear edge features. We integrate the recently developed mathematical construction of shearlets, which is more effective at identifying sparse anisotropic edges, with an existing automatic wavelet-based registration algorithm. Our shearlet features algorithm produces more distinct features than wavelet features algorithms; the separation of edges from textures is even stronger than with wavelets. Our algorithm computes shearlet and wavelet features for the images to be registered, then performs least squares minimization on these features to compute a registration transformation. Our algorithm is two-staged and multiresolution in nature. First, a cascade of shearlet features is used to provide a robust, though approximate, registration. This is then refined by registering with a cascade of wavelet features. Experiments across a variety of image classes show an improved robustness to initial conditions, when compared to wavelet features alone.

Better estimation of land surface microwave emissivity promises to improve over-land precipitation retrievals in the GPM era. Forward models of land microwave emissivity are available but have suffered from poor parameter specification and limited testing. Here, forward models are calibrated and the accompanying change in predictive power is evaluated. With inputs (e.g., soil moisture) from the Noah land surface model and applying MODIS LAI data, two microwave emissivity models are tested, the Community Radiative Transfer Model (CRTM) and Community Microwave Emission Model (CMEM). The calibration is conducted with the NASA Land Information System (LIS) parameter estimation subsystem using AMSR-E based emissivity retrievals for the calibration dataset. The extent of agreement between the modeled and retrieved estimates is evaluated using the AMSR-E retrievals for a separate 7-year validation period. Results indicate that calibration can significantly improve the agreement, simulating emissivity with an across-channel average root-mean-square-difference (RMSD) of about 0.013, or about 20% lower than if relying on daily estimates based on climatology. The results also indicate that calibration of the microwave emissivity model alone, as was done in prior studies, results in as much as 12% higher across-channel average RMSD, as compared to joint calibration of the land surface and microwave emissivity models. It remains as future work to assess the extent to which the improvements in emissivity estimation translate into improvements in precipitation retrieval accuracy.

Searching for relevant knowledge across heterogeneous geospatial databases requires an extensive knowledge of the semantic meaning of images, a keen eye for visual patterns, and efficient strategies for collecting and analyzing data with minimal human intervention. In this paper, we present our recently developed content-based multimodal Geospatial Information Retrieval and Indexing System (GeoIRIS) which includes automatic feature extraction, visual content mining from large-scale image databases, and high-dimensional database indexing for fast retrieval. Using these underpinnings, we have developed techniques for complex queries that merge information from heterogeneous geospatial databases, retrievals of objects based on shape and visual characteristics, analysis of multiobject relationships for the retrieval of objects in specific spatial configurations, and semantic models to link low-level image features with high-level visual descriptors. GeoIRIS brings this diverse set of technologies together into a coherent system with an aim of allowing image analysts to more rapidly identify relevant imagery. GeoIRIS is able to answer analysts' questions in seconds, such as "given a query image, show me database satellite images that have similar objects and spatial relationship that are within a certain radius of a landmark."