Surface Temperature Errors Research Articles

The impact of North Atlantic sea surface temperature errors on the simulation of North Atlantic European region climate

AbstractCurrent state‐of‐the‐art climate models fail to capture accurately the path of the Gulf Stream and North Atlantic Current. This leads to a warm bias near the North American coast, where the modelled Gulf Stream separates from the coast further north, and a cold anomaly to the east of the Grand Banks of Newfoundland, where the North Atlantic Current remains too zonal in this region. Using an atmosphere‐only model forced with the sea surface temperature (SST) biases in the North Atlantic, we consider the impact they have on the mean state and the variability in the North Atlantic European region in winter. Our results show that the SST errors produce a mean sea‐level pressure response that is similar in magnitude and pattern to the atmospheric circulation errors in the coupled climate model. The work also suggests that errors in the coupled model storm tracks and North Atlantic Oscillation, compared to reanalysis data, can also be explained partly by these SST errors. Our results suggest that both the error in the Gulf Stream separation location and the path of the North Atlantic Current around the Grand Banks play important roles in affecting the atmospheric circulation. Reducing these coupled model errors could improve significantly the representation of the large‐scale atmospheric circulation of the North Atlantic and European region. Copyright © 2012 Royal Meteorological Society

Does Model Parameter Error Cause a Significant “Spring Predictability Barrier” for El Niño Events in the Zebiak–Cane Model?

Abstract Within the framework of the Zebiak–Cane model, the approach of conditional nonlinear optimal perturbation (CNOP) is used to study the effect of model parameter errors on El Niño–Southern Oscillation (ENSO) predictability. The optimal model parameter errors are obtained within a reasonable error bound (i.e., CNOP-P errors), which have the largest effect on the results of El Niño predictions. The resultant prediction errors were investigated in depth. The CNOP-P errors do not cause a noticeable prediction error of the sea surface temperature anomaly averaged over the Niño-3 region and do not show an obvious season-dependent evolution of the prediction errors. Consequently, the CNOP-P errors do not cause a significant spring predictability barrier (SPB) for El Niño events. In contrast, the initial errors that have the largest effect on the results of the predictions (i.e., the CNOP-I errors) show a season-dependent evolution, with the largest error increase in spring, and also cause a large prediction error, thereby generating a significant SPB. The initial errors play a more important role than the parameter errors in generating a significant SPB for El Niño events. To further validate this result, the authors investigated the situation in which CNOP-I and CNOP-P errors are simultaneously superimposed in the model, which may be a more credible approach because the initial errors and model parameter errors coexist under realistic predictions. The combined mode of CNOP-I and CNOP-P errors shows a similar season-dependent evolution to that of CNOP-I errors and yields a large prediction error, thereby inducing a significant SPB. The inference, therefore, is that initial errors play a more important role than model parameter errors in generating a significant SPB for El Niño predictions of the Zebiak–Cane model. This result helps to clarify the roles of the initial error and parameter error in the development of an SPB, and highlights the role of initial errors, which demonstrates that the SPB could be markedly reduced by improving the initial conditions. The results provide a theoretical basis for improving data assimilation in ENSO predictions.

Error Sources in Remote Sensing of Microwave Land Surface Emissivity

The retrieval of land surface emissivity from satellite passive microwave measurements often requires the knowledge of various radiative components (e.g., atmospheric upwelling and downwelling radiation) contributed to the measurements. Under a cloud-free condition, atmospheric and surface radiative components can be derived from atmospheric temperature and water vapor, and surface temperature data. Thus, the quality of these auxiliary data sets directly affects the emissivity accuracy. From an emission-based radiative transfer equation, a set of relationships is derived to study the sensitivity of surface emissivity to the errors of brightness temperature, atmospheric transmittance, and surface temperature. As an example, the uncertainties in the Advanced Microwave Scanning Radiometer-Earth Observing System emissivity at 23 and 89 GHz may be much larger than the uncertainties of emissivity at lower frequencies due to the higher uncertainties in computing the water vapor absorption. The error in the land surface temperature is a main source of error in emissivity at the frequencies less than 19 GHz.

Estimation of instantaneous air temperature above vegetation and soil surfaces from Landsat 7 ETM+ data in northern Germany

The temperature–vegetation index method (TVX method, also called contextual method) for the area-wide mapping of instantaneous air temperature is adopted for use with Landsat 7 Enhanced Thematic Mapper Plus (ETM+) data. The method requires multispectral data consisting of bands in the red, near-infrared and thermal spectral regions, and no additional data. The approach is complemented with an iterative filtering routine for eliminating outliers and an interpolation algorithm for filling data gaps. The adopted method is applied to a multi-temporal dataset of nine ETM+ scenes, covering large parts of north-eastern Germany including the Durable Environmental Multidisciplinary Monitoring Information Network (DEMMIN) test site. Thus, for the first time the TVX method is applied to fine spatial resolution data and a central European region. The satellite-derived air temperatures (60 m spatial resolution) are compared with in situ measurements, showing an average error of about 3 K (root mean square, RMS), whereas the mean error in land surface temperature (LST) estimation is about 2 K. The results compare well with the in situ values throughout all seasons. The accuracy of about 3 K is in line with previously reported results for the TVX method (employing medium spatial resolution data) as well as for physically based approaches (ecosystem- or energy-balance models). Only remote sensing models incorporating in situ air temperature (as training data for neural networks or in multiple regression analysis) are reported to perform better in terms of RMS deviations. In the past, overestimation of air temperature by the TVX method was repeatedly observed. It is shown that the remote sensing approach tends to under- or overestimate the in situ air temperatures, depending on the in situ measurement heights. In conjunction with the attempt to assign the satellite-derived air temperature to a certain height above ground, the possibility of a simple correction for reference height is investigated. Over- and underestimations larger than 2 K seem to reflect existing differences in temperature rather than calculation errors. Furthermore, the dependence of the derived air temperature spatial pattern on different moving window sizes is shown. Possible sources of errors and limitations of the approach are discussed in detail.

High-resolution urban thermal sharpener (HUTS)

A high resolution urban thermal sharpener (HUTS) was developed that increases the resolution of thermal infrared (TIR) data to that of visible and near infrared (VNIR) data by fitting the relationship between radiometric surface temperature, normalized difference vegetation index ( NDVI) and surface albedo ( α). HUTS was applied to TIR data aggregated to 90 m to represent a satellite acquired dataset and validated against the measured 10 m data from an aircraft over San Juan, Puerto Rico. HUTS sharpening reduced the root mean square error of surface temperature at the high resolution by 17% compared to no sharpening and outperformed other sharpening methods. HUTS is proposed as a useful tool to study urban meteorology and climatology at the microscale using ASTER satellite data.

Use of carbon micro-particles for improved infrared temperature measurement of CMOS MEMS devices

We report a technique which can be used to improve the accuracy of infrared (IR) surface temperature measurements made on micro-electro-mechanical systems (MEMS). In this work, a silicon-on-insulator (SOI) complementary metal oxide semiconductor (CMOS) MEMS thermal flow sensor was studied. The device consists of a meander-shaped resistive heater element that was fabricated using CMOS aluminum metallization and embedded in an SOI silicon oxide (SiO2) membrane. Conventional IR temperature measurements, made on the active device, were shown to give significant surface temperature errors—a consequence of the optical transparency of the SiO2 membrane and low emissivity of the metalized areas. Radiative carbon micro-particles were used to improve the measurement accuracy. By making IR measurements on carbon micro-particles placed in isothermal contact with parts of the MEMS sensor's structure, the accuracy of the surface temperature determination was significantly improved. The peak device operating temperatures measured using the ‘IR micro-particle’ technique were found to be in good agreement with those obtained from thermoelectric characterization. The work shows how wide area surface temperature profiles can be quickly built up by imaging multiple micro-particles using an IR charge-coupled device (CCD) detector array. The IR micro-particle technique removes the problems associated with coating MEMS sensors with a uniform high emissivity layer, which may cause damage to such devices and/or affect their thermal performance due to additional heat spreading.

The radiation budget in a regional climate model

The long- and short-wave components of the radiation budget are among the most important quantities in climate modelling. In this study, we evaluated the radiation budget at the earth’s surface and at the top of atmosphere over Europe as simulated by the regional climate model CLM. This was done by comparisons with radiation budgets as computed by the GEWEX/SRB satellite-based product and as realised in the ECMWF re-analysis ERA40. Our comparisons show that CLM has a tendency to underestimate solar radiation at the surface and the energy loss by thermal emission. We found a clear statistical dependence of radiation budget imprecision on cloud cover and surface albedo uncertainties in the solar spectrum. In contrast to cloud fraction errors, surface temperature errors have a minor impact on radiation budget uncertainties in the long-wave spectrum. We also evaluated the impact of the number of atmospheric layers used in CLM simulations. CLM simulations with 32 layers perform better than do those with 20 layers in terms of the surface radiation budget components but not in terms of the outgoing long-wave radiation and of radiation divergence. Application of the evaluation approach to similar simulations with two additional regional climate models confirmed the results and showed the usefulness of the approach.

Errors analysis on temperature and emissivity determination from hyperspectral thermal infrared data

Sensitivity analysis of temperature-emissivity separation method commonly applied to hyperspectral data to various sources of errors is performed in this paper. In terms of resulting errors in the process of retrieving surface temperature, results show that: (1) Satisfactory results can be obtained for heterogeneous land surfaces and retrieval error of surface temperature is small enough to be neglected for all atmospheric conditions. (2) Separation of atmospheric downwelling radiance from at-ground radiance is not very sensitive to the uncertainty of column water vapor (WV) in the atmosphere. The errors in land surface temperature retrievals from at-ground radiance with the DRRI method due to the uncertainty in atmospheric downwelling radiance vary from -0.2 to 0.6K if the uncertainty of WV is within 50% of the actual WV; (3) Impact of the errors generated by the poor atmospheric corrections is significant, implying that a well-done atmospheric correction is indeed required to obtain accurate at-ground radiance from at-satellite radiance for successful separation of land-surface temperature and emissivity.

THE LAND SURFACE TEMPERATURE AND EMISSIVITY RETRIEVED FROM THE AMSR PASSIVE MICROWAVE DATA

AMSR and MODIS are two EOS (Earth Observing System) instruments in Aqua satellite. A regression analysis between the brightness of all AMSR bands and the MODIS land surface temperature product provided by NASA indicates good correlation, and hence retrieving land surface temperature from AMSR passive data is available without ground true data (soil moisture and land surface type). The analytical results in such regions as North Africa, Northeast China, Tibet China and India indicate that the radiation mechanism of snow-covered surface is different from mechanism of other conditions. In order to retrieve land surface temperature more accurately, we can classify the land surface at least into two groups. For land surface covered with no snow, the average land surface temperature error is about 2℃ relative to the MODIS LST product. For snow-covered land surface, the average land surface temperature error is about 2~3℃ relative to the MODIS LST product. Besides, the emissivity of passive microwave is a very important parameter for retrieving soil moisture. The authors computed the emissivity through land surface temperature retrieved by the statistical regression method and made an analysis.

AN OPTIMIZAED MULTIPLE-BAND ALGORITHM BY USING NEURAL NETWORK FOR SEPARATING LAND SURFACE EMISSIVITY AND TEMPERATURE FROM ASTER IMAGERY

A multiple-band algorithm is proposed in this paper to separate land surface temperature and emissivity from ASTER data.Three methods can be used to solve the equations.The first is the performance of classification for the images and the formulation of different equations,followed by the solution of the equations. The second is least-squares.The third is the simulation of the database according to the characteristics of object emissivities and the utilization of the neural network to solve equations.An analysis indicates that the neural network can improve the practicability and accuracy of the algorithm.The accuracy of neural network proves to be very high for the test data simulated from MODTRAN 4.An application example is given in this paper,and the analysis suggests that the neural network also possesses the self-study capability.The simulation data show that the average error of land surface temperature is below 0.5℃,and the error of emissivity in band 11~14 is below 0.007(band 11,12)and 0.006(band 13,14),respectively.

Validation of the North American ASTER Land Surface Emissivity Database (NAALSED) version 2.0 using pseudo-invariant sand dune sites

Knowledge of the Land Surface Emissivity (LSE) in the Thermal Infrared (TIR: 8–12 µm) part of the electromagnetic spectrum is essential to derive accurate Land Surface Temperatures (LSTs) from spaceborne TIR measurements. This study focuses on validation of the emissivity product in the North American ASTER Land Surface Emissivity Database (NAALSED) v2.0 — a mean seasonal, gridded emissivity product produced at 100 m spatial resolution using all Advanced Spaceborne Thermal Emission and Reflection Radiometer (ASTER) scenes from 2000 to 2008 over North America ( http://emissivity.jpl.nasa.gov). The NAALSED emissivity product was validated over bare surfaces with laboratory measurements of sand samples collected at nine pseudo-invariant sand dune sites located in the western/southwestern USA. The nine sand dune sites cover a broad range of surface emissivities in the TIR. Results show that the absolute mean emissivity difference between NAALSED and the laboratory results for the nine validation sites and all five ASTER TIR bands was 0.016 (1.6%). This emissivity difference is equivalent to approximately a 1 K error in the land surface temperature for a material at 300 K in the TIR.

Recent changes in the North Atlantic circulation simulated with eddy-permitting and eddy-resolving ocean models

Three eddy-permitting (1/4°) versions and one eddy-resolving (1/12°) version of the OCCAM ocean model are used to simulate the World Ocean circulation since 1985. The first eddy-permitting simulation has been used extensively in previous studies, and provides a point of reference. A second, improved, eddy-permitting simulation is forced in the same manner as the eddy-resolving simulation, with a dataset based on a blend of NCEP re-analysis and satellite data. The third eddy-permitting simulation is forced with a different dataset, based on the ERA-40 re-analysis data. Inter-comparison of these simulations in the North Atlantic clarifies the relative importance of resolution and choice of forcing dataset, for simulating the mean state and recent variability of the basin-scale circulation in that region. Differences between the first and second eddy-permitting simulations additionally reveal an erroneous influence of sea ice on surface salinity, dense water formation, and the meridional overturning circulation. Simulations are further evaluated in terms of long-term mean ocean heat transport at selected latitudes (for which hydrographic estimates are available) and sea surface temperature errors (relative to observations). By these criteria, closest agreement with observations is obtained for the eddy-resolving simulation. In this simulation, there is also a weak decadal variation in mid-latitudes, with heat transport strongest, by around 0.2 PW, in the mid-1990s. In two of the eddy-permitting simulations, by contrast, heat transport weakens through the study period, by up to 0.4 PW in mid-latitudes. The most notable changes of heat transport in all simulations are linked to a weakening of the subpolar gyre, rather than changes in the meridional overturning circulation. It is concluded that recent changes in the structure of mid-latitude heat transport in the North Atlantic are more accurately represented if eddies are explicitly resolved.

Calibrating and Evaluating Reanalysis Surface Temperature Error by Topographic Correction

AbstractBased on the observed daily surface air temperature data from 597 stations over continental China and two sets of reanalysis data [NCEP–NCAR and 40-yr ECMWF Re-Analysis (ERA-40)] during 1979–2001, the altitude effects in calibrating and evaluating reanalyzed surface temperature errors are studied. The results indicate that the accuracy of interpolated surface temperature from the reanalyzed gridpoint value or the station observations depends much on the altitudes of original data. Bias of interpolated temperature is usually in proportion to the increase of local elevation and topographical complexity. Notable improvements of interpolated surface temperature have been achieved through “topographic correction,” especially for ERA-40, which highlights the necessity of removal of “elevation-induced bias” when using and evaluating reanalyzed surface temperature.

A Neural Network Technique for Separating Land Surface Emissivity and Temperature From ASTER Imagery

Four radiative transfer equations for Advanced Spaceborne Thermal Emission and Reflection Radiometer (ASTER) bands 11, 12, 13, and 14 are built involving six unknowns (average atmospheric temperature, land surface temperature, and four band emissivities), which is a typical ill-posed problem. The extra equations can be built by using linear or nonlinear relationship between neighbor band emissivities because the emissivity of every land surface type is almost constant for bands 11, 12, 13, and 14. The neural network (NN) can make full use of potential information between band emissivities through training data because the NN simultaneously owns function approximation, classification, optimization computation, and self-study ability. The training database can be built through simulation by MODTRAN4 or can be obtained from the reliable measured data. The average accuracy of the land surface temperature is about 0.24 K, and the average accuracy of emissivity in bands 11, 12, 13, and 14 is under 0.005 for test data. The retrieval result by the NN is, on average, higher by about 0.7 K than the ASTER standard product (AST08), and the application and comparison indicated that the retrieval result is better than the ASTER standard data product. To further evaluate self-study of the NN, the ASTER standard products are assumed as measured data. After using AST09, AST08, and AST05 (ASTER Standard Data Product) as the compensating training data, the average relative error of the land surface temperature is under 0.1 K relative to the AST08 product, and the average relative error of the emissivity in bands 11, 12, 13, and 14 is under 0.001 relative to AST05, which indicates that the NN owns a powerful self-study ability and is capable of suiting more conditions if more reliable and high-accuracy ASTER standard products can be compensated.

Using numerical weather prediction errors to estimate aerosol heating

The total real response of the atmosphere to aerosols can be predicted by examining numerical models. In this study, the moderate resolution imaging spectroradiometer (MODIS) aerosol optical thickness (AOT) is correlated to model temperature errors to estimate this response for Israel and for Italy. Significant correlations between aerosols and atmospheric numerical model temperature errors are presented. Two main results were found in this study. First, the correlation between the UK Metrological Office (UKMO) model temperature error (ΔT) at level 850 hPa for Tel Aviv (Israel) during the year 2002 and the MODIS AOT ≥ 0.5, was found to be −0.54. Second, the sign of the correlation between the European Center for Medium-RangeWeather Forecasting (ECMWF) monthly averaged model surface temperature errors for Italy and the MODIS AOT depends on the aerosol type, as reflected by the specific area. The correlation between air surface ECMWF ΔT and AOT was found to be +0.70 in southern Italy. Although the contribution of aerosols is ignored by most models, the findings presented here are statistically significant and provide a first reliable estimation of the realistic full atmospheric temperature response to aerosol processes.

A physics-based statistical algorithm for retrieving land surface temperature from AMSR-E passive microwave data

AMSR-E and MODIS are two EOS (Earth Observing System) instruments on board the Aqua satellite. A regression analysis between the brightness of all AMSR-E bands and the MODIS land surface temperature product indicated that the 89 GHz vertical polarization is the best single band to retrieve land surface temperature. According to simulation analysis with AIEM, the difference of different frequencies can eliminate the influence of water in soil and atmosphere, and also the surface roughness partly. The analysis results indicate that the radiation mechanism of surface covered snow is different from others. In order to retrieve land surface temperature more accurately, the land surface should be at least classified into three types: water covered surface, snow covered surface, and non-water and non-snow covered land surface. In order to improve the practicality and accuracy of the algorithm, we built different equations for different ranges of temperature. The average land surface temperature error is about 2–3°C relative to the MODIS LST product.

Effects of tropical sea surface temperature (SST) errors on the Antarctic atmospheric circulation of HadCM3

This paper investigate the large mean sea level pressure errors in the HadCM3 atmosphere‐ocean global climate model around Antarctica and finds them to be caused by the sea surface temperature (SST) errors in the winter tropics. This teleconnection is explained by the warm tropical SST anomalies causing anomalous ascent which strengthens the Hadley circulation and increases the high‐level (∼200hPa) divergence. This in turn interacts with the strong meridional gradient in absolute vorticity to generate Rossby waves that propagate polewards. By imposing SST anomalies, in the atmosphere only model, in different ocean basins it is shown that it is the SST errors over Indonesia and the eastern tropical Atlantic that are responsible for these wave trains; and hence, primarily responsible for the pressure anomalies. By contrast, a large SST error over the eastern tropical Pacific does not cause a wave train to be generated because it does not generate strong ascent. This study also demonstrates the difficulties in diagnosing errors within a coupled climate model since forcing, particularly in the tropics, may have a strongly non‐local effect.

An intercomparison of the Surface Energy Balance Algorithm for Land (SEBAL) and the Two-Source Energy Balance (TSEB) modeling schemes

An intercomparison of output from two models estimating spatially distributed surface energy fluxes from remotely sensed imagery is conducted. A major difference between the two models is whether the soil and vegetation components of the scene are treated separately (Two-Source Energy Balance; TSEB approach) or as a lumped composite (one-source approach; Surface Energy Balance Algorithm for Land; SEBAL) in the parameterization of radiative and turbulent exchanges with the overlying air. Comparisons are performed using data from two largescale field experiments covering sub-humid grassland (Southern Great Plains '97) and semi-arid rangeland (Monsoon '90) having very different landscape properties. In general, there was reasonable agreement between flux output from both models versus a handful of flux tower observations. However, spatial intercomparisons of model output over the full modeling domains yielded relatively large discrepancies (on the order of 100 W m − 2 ) in sensible heat flux ( H) that are related to land cover. In particular, bare soil and sparsely vegetated areas yielded the largest discrepancies, with TSEB fluxes being in better agreement with tower observations. Modifications to SEBAL inputs that reduced discrepancies with TSEB and observations for bare soil and shrub classes tended to increase differences for other land cover classes. In particular, improvements to SEBAL inputs of surface roughness for momentum tended to exacerbate errors with respect to observed fluxes. These results suggest that some of the simplifying assumptions in SEBAL may not be strictly applicable over the wide range in conditions present within these landscapes. An analysis of TSEB and SEBAL sensitivity to uncertainties in primary inputs indicated that errors in surface temperature or surface-air temperature differences had the greatest impact on H estimates. Inputs of secondary importance were fractional vegetation cover for TSEB, while for SEBAL, the selection of pixels representing wet and dry moisture end-member conditions significantly influenced flux predictions. The models were also run using input fields derived from both local and remote data sources, to test performance under conditions of varying ancillary data availability. In this case, both models performed similarly under both constraints.

Regional evaporation estimates from flux tower and MODIS satellite data

Two models were evaluated for their ability to estimate land surface evaporation at 16-day intervals using MODIS remote sensing data and surface meteorology as inputs. The first was the aerodynamic resistance–surface energy balance model, and the second was the Penman–Monteith (P–M) equation, where the required surface conductance is estimated from remotely-sensed leaf area index. The models were tested using 3 years of evaporation and meteorological measurements from two contrasting Australian ecosystems, a cool temperate, evergreen Eucalyptus forest and a wet/dry, tropical savanna. The aerodynamic resistance–surface energy balance approach failed because small errors in the radiative surface temperature translate into large errors in sensible heat, and hence into estimates of evaporation. The P–M model adequately estimated the magnitude and seasonal variation in evaporation in both ecosystems (RMSE = 27 W m − 2 , R 2 = 0.74), demonstrating the validity of the proposed surface conductance algorithm. This, and the ability to constrain evaporation estimates via the energy balance, demonstrates the superiority of the P–M equation over the surface temperature-based model. There was no degradation in the performance of the P–M model when gridded meteorological data at coarser spatial (0.05°) and temporal (daily) resolution were substituted for locally-measured inputs. The P–M approach was used to generate a monthly evaporation climatology for Australia from 2001 to 2004 to demonstrate the potential of this approach for monitoring land surface evaporation and constructing monthly water budgets from 1-km to continental spatial scales.

Predictability Loss in an Intermediate ENSO Model due to Initial Error and Atmospheric Noise*

AbstractThe seasonal and interannual predictability of ENSO variability in a version of the Zebiak–Cane coupled model is examined in a perturbation experiment. Instead of assuming that the model is “perfect,” it is assumed that a set of optimal initial conditions exists for the model. These states, obtained through a nonlinear minimization of the misfit between model trajectories and the observations, initiate model forecasts that correlate well with the observations. Realistic estimates of the observational error magnitudes and covariance structures of sea surface temperatures, zonal wind stress, and thermocline depth are used to generate ensembles of perturbations around these optimal initial states, and the error growth is examined. The error growth in response to subseasonal stochastic wind forcing is presented for comparison.In general, from 1975 to 2002, the large-scale uncertainty in initial conditions leads to larger error growth than continuous stochastic forcing of the zonal wind stress fields. Forecast ensemble spread is shown to depend most on the calendar month at the end of the forecast rather than the initialization month, with the seasons of greatest spread corresponding to the seasons of greatest anomaly variance. It is also demonstrated that during years with negative (and rapidly decaying) Niño-3 SST anomalies (such as the time period following an El Niño event), there is a suppression of error growth. In years with large warm ENSO events, the ensemble spread is no larger than in moderately warm years. As a result, periods with high ENSO variance have greater potential prediction utility.In the realistic range of observational error, the ensemble spread has more sensitivity to the initial error in the thermocline depth than to the sea surface temperature or wind stress errors. The thermocline depth uncertainty is the principal reason why initial condition uncertainties are more important than wind noise for ensemble spread.