Relative Humidity Errors Research Articles

Adaptive tuning of uncertain parameters in a numerical weather prediction model based upon data assimilation

AbstractIn numerical weather prediction models, near‐surface quantities like 10‐m wind speed (FF10M) or 2‐m temperature (T2M) tend to exhibit significantly larger forecast errors than the related variables in the free troposphere. Besides representativeness issues of surface stations, this is primarily related to parametrization errors and insufficient knowledge of relevant physical properties of the soil and the surface layer. For instance, the vegetation roughness length is usually derived from land‐cover classifications that may contain errors and reflect only part of the natural variability. This article describes a methodology implemented into Deutscher Wetterdienst's operational numerical weather prediction model ICON in order to improve the estimate of such parameters by using information from the data assimilation system. Building upon the condition that FF10M, T2M, and 2‐m relative humidity are assimilated, time‐filtered assimilation increments are calculated for the respective fields at the lowest model level. These are taken as proxies for the related model biases. For T2M, an additional weighted increment field is computed that indicates the model bias in the diurnal temperature amplitude. Based on these increment fields, several physical parameter fields and a few model tuning parameters are varied around their base values. This adaptive parameter adjustment is used operationally in the global and regional forecasting systems of Deutscher Wetterdienst. The ensuing reduction of the FF10M, T2M, and 2‐m relative humidity errors typically lies on the order of 5% on a hemispheric average but has substantial regional and seasonal variability that depends on the original magnitude of the model error. A weaker but still statistically significant positive impact is seen in the radiosonde verification of wind speed, humidity, and temperature in the lower troposphere, giving confidence that the adaptive tuning indeed reduces model errors rather than pushing the model towards unrepresentative station observations.

Neuro-fuzzy humidity control for white oyster mushroom in a closed plant production system

Oyster mushrooms grow optimally in relative humidity of 80-90%. As a tropical country, Indonesia is very suitable for oyster mushroom cultivation, but relative humidity must be controlled. The usual effort is ineffective by spraying water to mushroom inside kumbung (mushroom cultivation building) every morning and evening. A building called a Closed Plant Production System (CPPS) equipped with a control system needs to be developed as a solution. The purpose of this study was to design and do a performance test of a neuro-fuzzy based relative humidity control system inside CPPS for white oyster mushroom (Pleurotus ostreatus) cultivation. The main components consist of Arduino Mega 2560 microcontroller, SHT11 sensor, and diaphragm pump connects with mist nozzle as an actuator. The performance test was carried out by running neuro-fuzzy based control system and giving disturbance inside CPPS room. As a result, this control system was able to reach a steady-state condition within 5 minutes. The highest relative humidity error was 0.73%, and the lowest error was 0.09%—the recovery rate of relative humidity inside CPPS when disturbed was 2.25% per minute.

Estimating Error Variances of a Microwave Sensor and Dropsondes aboard the Global Hawk in Hurricanes Using the Three-Cornered Hat Method

AbstractThis study estimates the random error variances and standard deviations (STDs) for four datasets: Global Hawk (GH) dropsondes (DROP), the High-Altitude Monolithic Microwave Integrated Circuit Sounding Radiometer (HAMSR) aboard the GH, the fifth European Centre for Medium-Range Weather Forecasts (ECMWF) reanalysis (ERA5), and the Hurricane Weather Research and Forecasting (HWRF) Model, using the three-cornered hat (3CH) method. These estimates are made during the 2016 Sensing Hazards with Operational Unmanned Technology (SHOUT) season in the environment of four tropical cyclones from August to October. For temperature and specific and relative humidity, the ERA5, HWRF, and DROP datasets all have similar magnitudes of errors, with ERA5 having the smallest. The error STDs of temperature and specific humidity are less than 0.8 K and 1.0 g kg−1 over most of the troposphere, while relative humidity error STDs increase from less than 5% near the surface to between 10% and 20% in the upper troposphere. The HAMSR bias-corrected data have larger errors, with estimated error STDs of temperature and specific humidity in the lower troposphere between 1.5 and 2.0 K and between 1.5 and 2.5 g kg−1. HAMSR’s relative humidity error STD increases from approximately 10% in the lower troposphere to 30% in the upper troposphere. The 3CH method error estimates are generally consistent with prior independent estimates of errors and uncertainties for the HAMSR and dropsonde datasets, although they are somewhat larger, likely due to the inclusion of representativeness errors (differences associated with different spatial and temporal scales represented by the data).

Research on new water-cooled textile air conditioning system

Aiming at the existing textile air conditioning system in China, the temperature and relative humidity do not meet the process requirements, the air exchange is uneven, and the energy consumption is high.This paper refers to the working principle of traditional textile air conditioner, combined with the requirements of temperature and humidity in textile workshop, and designs a new water-cooled textile air conditioning system from three aspects: air conditioning design, various components and Flow simulation.The results show that the temperature and relative humidity errors are 4.3% and 2.0%, respectively. The spring temperature can be maintained at 22.1 °C, the relative humidity is 77.3%, the electric energy is saved by 30%, and the water is savedby 27%.

Error analysis of upper tropospheric water vapor in CMIP5 models using “A-Train” satellite observations and reanalysis data

Upper tropospheric water vapor (UTWV) plays a critical role in amplifying global warming caused by increasing greenhouse gases, yet it is one of the most poorly simulated quantities in climate models. It is not clear what physical processes play a central role in controlling the model errors in UTWV. We diagnose the UTWV simulation errors from AMIP models submitted to the CMIP5 project by using “A-Train” satellite observation and reanalysis data. We identify the relative contributions of errors in relative humidity (RH), temperature, and large-scale circulation (represented by vertical pressure velocity at 500 hPa, ω500) to the modeled UTWV errors over the tropics (30°N–30°S). It is found that models generally have positive biases in UTWV, except over the continental convective regions where negative biases predominate. The errors in the patterns and amplitudes of climatological UTWV are highly correlated with those in RH and ω500. The fractional UTWV errors show large positive errors over the large-scale descending regimes (0 300 K or ω500 < −30 hPa/day) and these errors are positive (negative) where anomalous descent (ascent) occurs during El Nino. We find that the water vapor errors are dominated by the errors in RH rather than in temperature throughout the troposphere, while temperature errors play an important role for water vapor errors near the tropopause.

Wintertime Subkilometer Numerical Forecasts of Near-Surface Variables in the Canadian Rocky Mountains

Abstract Numerical weather prediction (NWP) systems operational at many national centers are nowadays used at the kilometer scale. The next generation of NWP models will provide forecasts at the subkilometer scale. Large impacts are expected in mountainous terrain characterized by highly variable orography. This study investigates the ability of the Canadian NWP system to provide an accurate forecast of near-surface variables at the subkilometer scale in the Canadian Rocky Mountains in wintertime when the region is fully covered by snow. Observations collected at valley and high-altitude stations are used to evaluate forecast accuracy at three different grid spacing (2.5, 1, and 0.25 km) over a period of 15 days. Decreasing grid spacing was found to improve temperature forecasts at high-altitude stations because of better orography representation. In contrast, no improvement is obtained at valley stations due to an inability of the model to fully capture at all resolutions the intensity of valley cold pools forming during nighttime. Errors in relative humidity reveal that the model tends to overestimate relative humidity at all resolutions, without improvement with decreasing grid spacing. Wind speed forecasts show large improvements with decreasing grid spacing for high-altitude stations exposed to or sheltered from wind. However, no systematic improvement with decreasing grid spacing is found for all stations, which is similar to previous studies. In addition, the model’s sensitivity at subkilometer grid spacing is investigated by evaluating the effects of (i) accounting for additional drag generated by subgrid orographic features, (ii) considering slope angle and aspect on surface radiation, and (iii) using high-resolution initialization for the surface fields.

Fluid dynamic analysis on solar heating error of radiosonde humidity measurement

The error of humidity sensor induced by solar radiation seriously affects the accuracy of the relative humidity measurement. To solve this problem, this paper presents a novel numerical analysis method of correcting the error of relative humidity based on computational fluid dynamics. In view of the external thermal environmental conditions of radiosonde humidity sensors, the convection-solar radiation coupled thermal boundary conditions are adopted in the numerical simulation with analysis method of fluid-solid coupled heat transfer. The temperature error analysis model is first established from the ground to 32 km altitude with different pressures and temperatures. Combined with Goff-Gratch approximation formulas of saturation vapor pressure, the corresponding fluid dynamic numerical analysis model of relative humidity is put forward for error correction. Moreover, the relative humidity errors are reported in the different physical parameters such as the direction of the solar radiation, the reflectivity of sensor, thermal conductivity of the substrate material, the size of sensor, etc. The data analysis shows that the error of relative humidity, caused by solar radiation nonlinearly increases with altitude. The humidity measurement accuracy is affected notably by the direction of solar radiation. Among the errors caused by solar radiation, the error in the direction perpendicular to the front of the sensor is biggest, the error in the direction of the top of the sensor is the next, and the error in the direction of the side of the sensor is smallest. The data analysis also indicates that the solar heating error of the relative humidity can be reduced by reducing the size of the sensor, reducing the thermal conductivity of the substrate material, or improving the reflectivity of sensor. However, the solar heating error can not be neglected under low air pressure at high altitude. A comparison with experimental results shows that the numerical analysis method of the relative humidity error based on fluid dynamics simulation provides a new way to enhance the radiation error correction.

Verification of RUC 0–1-h Forecasts and SPC Mesoscale Analyses Using VORTEX2 Soundings

Abstract This study uses radiosonde observations obtained during the second phase of the Verification of the Origins of Rotation in Tornadoes Experiment (VORTEX2) to verify base-state variables and severe-weather-related parameters calculated from Rapid Update Cycle (RUC) analyses and 1-h forecasts, as well as those calculated from the operational surface objective analysis system used at the Storm Prediction Center (the SFCOA). The rapid growth in temperature, humidity, and wind errors from 0 to 1 h seen at all levels in a past RUC verification study by Benjamin et al. is not seen in the present study. This could be because the verification observations are also assimilated into the RUC in the Benjamin et al. study, whereas the verification observations in the present study are not. In the upper troposphere, the present study shows large errors in relative humidity, mostly related to a large moist bias. The planetary boundary layer tends to be too shallow in the RUC analyses and 1-h forecasts. Wind speeds tend to be too fast in the lowest 1 km and too slow in the 2–4-km layer. RUC and SFCOA 1-h forecast errors for many important severe weather parameters are large relative to their potential impact on convective evolution. However, the SFCOA significantly improves upon the biases seen in most of the 1-h RUC forecasts for the base-state surface variables and most of the other severe-weather-related parameters, indicating that the SFCOA has a more significant impact in reducing the biases in the 1-h RUC forecasts than on the root-mean-squared errors.

Assessing the performance of a prognostic and a diagnostic cloud scheme using single column model simulations of TWP–ICE

AbstractSingle column model simulations using the UK Met Office Unified Model, as used in the Australian Community Climate Earth System Simulator, are presented for the Tropical Warm Pool–International Cloud Experiment (TWP–ICE) field study. Two formulations for the representation of clouds are compared with the extensive observations taken during the campaign, giving insight into the ability of the model to simulate tropical cloud systems. During the active monsoon phase the modelled cloud cover has a stronger dependence on relative humidity than the observations. Observed ice cloud properties during the suppressed monsoon period show that the ice water content is significantly underestimated in the simulations. The profiles of modelled ice fall speeds are faster than those observed in the levels above 12 km, implying that the observations have smaller sized particles in larger concentrations than the models. Both simulations show similar errors in the diurnal cycle of relative humidity during the active monsoon phase, suggesting that the error is less sensitive to the choice of cloud scheme and rather is driven by the convection scheme. However, during the times of suppressed convection the relative humidity error is different between the simulations, with congestus convection drying the environment too much, particularly in the prognostic cloud‐scheme simulation. This result shows that the choice of cloud scheme and the way that the cloud and convection schemes interact plays a role in the temperature and moisture errors during the suppressed monsoon phase, which will impact the three‐dimensional model simulations of tropical variability. Copyright © 2011 Royal Meteorological Society

Relative humidity errors when measuring dry and wet bulb temperatures

Relative humidity ( h r ) errors, implied in the psychrometric method of measurement, were calculated as a function of the accuracies in temperature sensors, regardless of their origin. An error propagation formula was used which included partial derivatives of h r with respect to dry ( T a ) and wet ( T wb ) bulb temperatures. Relative humidity errors ( E hr ) were predicted for T a between 5 and 35 °C, percentage h r from 30 to 99% and atmospheric pressures from 0.6 x 10 5 to 1 × 10 5 Pa. Sensor accuracies from 0.1 to 1 °C were considered. The latter value, when occurring simultaneously in both sensors, led to an E hr value as large as 9% (in percentage relative humidity units). For dry bulb temperatures varying from 15 to 25 °C and percentage h r from 50 to 80%, a sensor accuracy of 0.2 °C in dry and wet bulb temperature sensors led to E hr at or below 2%. Inexpensive temperature sensors with accuracies higher than 0.5 °C lead to unacceptably high errors in relative humidity. Values of E hr increase linearly with the increase in relative humidity and decrease exponentially with the increase in dry bulb temperature. The reduction of atmospheric pressure causes a very slight linear decrease in the relative humidity errors.

Sensitivity analysis of a Penman–Monteith type equation to estimate reference evapotranspiration in southern Spain

AbstractSensitivity analysis is crucial in assessing the impact of climatic variables on reference evapotranspiration estimations. The sensitivity of the standardized ASCE–Penman–Monteith evapotranspiration equation for daily estimations to climatic variables has not yet been studied in Spain. Andalusia is located in southern Spain where almost 1 million ha are irrigated under quite different conditions; it has a high inter‐annual variability in rainfall. In this study, sensitivity analyses for this equation were carried out for temperature, relative humidity, solar radiation and wind speed data from 87 automatic weather stations, including coastal and inland locations, from 1999 to 2006. Topography and Mediterranean climate characterize the heterogeneous landscape and vegetation of this region. Simulated random and systematic errors have been added to meteorological data to obtain ET0 deviations and sensitivity coefficients for different time periods. BIAS and SEE (standard error of estimate) have been used to evaluate the effect of both types of errors. The results showed a large degree of daily and seasonal variability, especially for temperature and relative humidity. In general, the effect on ET0 values of introduced random errors was larger than that of systematic errors. ET0 overestimations were produced using positive errors in temperature, solar radiation and wind speed data, while these errors in relative humidity resulted in ET0 underestimations. The sensitivity of ET0 to the same climatic variables showed significant differences among locations. The geographical distribution of sensitivity coefficients across this region was also studied. As an example, during spring months, ET0 equation was more sensitive to temperature in stations located along the Guadalquivir Valley. Copyright © 2009 John Wiley &amp; Sons, Ltd.

Dry Bias in Vaisala RS90 Radiosonde Humidity Profiles over Antarctica

Abstract Middle to upper tropospheric humidity plays a large role in determining terrestrial outgoing longwave radiation. Much work has gone into improving the accuracy of humidity measurements made by radiosondes. Some radiosonde humidity sensors experience a dry bias caused by solar heating. During the austral summers of 2002/03 and 2003/04 at Dome C, Antarctica, Vaisala RS90 radiosondes were launched in clear skies at solar zenith angles (SZAs) near 83° and 62°. As part of this field experiment, the Polar Atmospheric Emitted Radiance Interferometer (PAERI) measured downwelling spectral infrared radiance. The radiosonde humidity profiles are used in the simulation of the downwelling radiances. The radiosonde dry bias is then determined by scaling the humidity profile with a height-independent factor to obtain the best agreement between the measured and simulated radiances in microwindows between strong water vapor lines from 530 to 560 cm−1 and near line centers from 1100 to 1300 cm−1. The dry biases, as relative errors in relative humidity, are 8% ± 5% (microwindows; 1σ) and 9% ± 3% (line centers) for SZAs near 83°; they are 20% ± 6% and 24% ± 5% for SZAs near 62°. Assuming solar heating is minimal at SZAs near 83°, the authors remove errors that are unrelated to solar heating and find the solar-radiation dry bias of 9 RS90 radiosondes at SZAs near 62° to be 12% ± 6% (microwindows) and 15% ± 5% (line centers). Systematic errors in the correction are estimated to be 3% and 2% for microwindows and line centers, respectively. These corrections apply to atmospheric pressures between 650 and 200 mb.

Evolution of ozone, particulates, and aerosol direct radiative forcing in the vicinity of Houston using a fully coupled meteorology‐chemistry‐aerosol model

A new fully coupled meteorology‐chemistry‐aerosol model is used to simulate the urban‐ to regional‐scale variations in trace gases, particulates, and aerosol direct radiative forcing in the vicinity of Houston over a 5 day summer period. Model performance is evaluated using a wide range of meteorological, chemistry, and particulate measurements obtained during the 2000 Texas Air Quality Study. The predicted trace gas and particulate distributions were qualitatively similar to the surface and aircraft measurements with considerable spatial variations resulting from urban, power plant, and industrial sources of primary pollutants. Sulfate, organic carbon, and other inorganics were the largest constituents of the predicted particulates. The predicted shortwave radiation was 30 to 40 W m−2 closer to the observations when the aerosol optical properties were incorporated into the shortwave radiation scheme; however, the predicted hourly aerosol radiative forcing was still underestimated by 10 to 50 W m−2. The predicted aerosol radiative forcing was larger over Houston and the industrial ship channel than over the rural areas, consistent with surface measurements. The differences between the observed and simulated aerosol radiative forcing resulted from transport errors, relative humidity errors in the upper convective boundary layer that affect aerosol water content, secondary organic aerosols that were not yet included in the model, and uncertainties in the primary particulate emission rates. The current model was run in a predictive mode and demonstrates the challenges of accurately simulating all of the meteorological, chemical, and aerosol parameters over urban to regional scales that can affect aerosol radiative forcing.

An Evaluation of Mesoscale-Model-Based Model Output Statistics (MOS) during the 2002 Olympic and Paralympic Winter Games

Abstract The skill of a mesoscale-model-based model output statistics (MOS) system that provided hourly forecasts for 18 sites over northern Utah during the 2002 Winter Olympic and Paralympic Games is evaluated. The MOS system was developed using three winters (November–April 1998/99, 1999/2000, and 2000/01) of forecasts by the fifth-generation Pennsylvania State University–National Center for Atmospheric Research Mesoscale Model (MM5) and observations from Olympic venues and transportation corridors. MOS temperature, relative humidity, wind speed, and wind direction forecasts were considerably more accurate than those produced by the 12- and 4-km MM5 grids. A primary contributor to MM5 temperature and relative humidity errors was a systematic overprediction of surface temperature (i.e., a warm/dry bias) during persistent and nocturnal cold-pool events when corresponding errors in MM5 dewpoint temperature forecasts were not observed. MOS largely corrected for this temperature bias. MOS wind speed forecast...

Retrieval of atmospheric parameters from NOAA-16 AMSU data over Indian region - Preliminary results

With the establishment of a new High Resolution Picture Transmission (HRPT) reception system at IMD, New Delhi, the real time data from microwave sounding instruments onboard the NOAA-K, L, M and N series of satellites has also become available. The raw HRPT data is being interfaced with the recently acquired new ‘ATOVS and AVHRR Preprocessing Package (AAPP)’ to perform temperature and moisture retrievals from AMSU data of NOAA-16 satellite using two separate schemes: Inversion Coupled Imager (ICI) and Neural Network (NN) approach. In this study, NOAA-16 satellite data over Indian region were used for retrieving temperature and moisture profiles for the month of January, 2002. The temperature and moisture retrieval results are evaluated by computing the bias and root mean square (RMS) difference using collocated ECMWF analysis. The results based on the analysis of data set for the month of January, 2002 shows that ICI approach yields better results for all atmospheric levels. The RMS errors in temperature profiles are found to be less than 4oC at all pressure levels. The RMS errors in relative humidity are found to be less than 20% at all pressure levels. C Intercomparison of the results revealed that bias and RMS error are less for ICI scheme using forecast field compared to ICI without using forecast field and Neural Network approach.

Vertical Correlation Functions for Temperature and Relative Humidity Errors

This article gives the details and results of an investigation into the properties of the temperature and relative humidity errors from the Navy Operational Global Atmospheric Prediction System for a 4-month period from March to June 1998. The spatial covariance data for temperature errors and for relative humidity errors were fit using eight different approximation functions/weighting methods. From these, two were chosen as giving good estimates of the parameters and variances of the prediction and observation errors and were used in further investigations. The vertical correlation between temperature errors at different levels and relative humidity errors at different levels was approximated using a combination of functional fitting and transformation of the pressure levels. The cross covariance between temperature and relative humidity errors at various pressure levels were approximated in two ways: 1) by directly computing and approximating the cross-covariance data, and 2) by approximating variance of the difference of normalized data. The latter led to more consistent results. Figures illustrating the results are included.

Soil‐moisture nudging experiments with a single‐column version of the ECMWF model

AbstractThe soil‐moisture nudging technique suggests using model forecast errors in near‐surface air temperature and relative humidity to re‐initialize (update) soil moisture in atmospheric models. This study investigates the application of soil‐moisture nudging using a single‐column version of the European Centre for Medium‐Range Weather Forecasts (ECMWF) model. the model was applied at 16 sites selected to sample a range of climates and land covers across the globe, with atmospheric forcing taken from the ECMWF operational analysis for 15 June 1994 and 15 December 1994. When observation errors are set to zero, the Optimal Interpolation technique for deriving estimates of nudging coefficients shows computational instability because strong correlation between near‐surface temperature and near‐surface relative humidity forecast errors makes the coefficient matrix of the linear equations ill‐posed. Therefore, Principal Component Analysis (PCA) is used as a pre‐processor to identify the independent and dependent components of temperature and relative humidity errors. With PCA, the soilmoisture nudging coefficients appropriate to the dominant principal component become stable and effective for correcting soil moisture within days. Such coefficients, when derived for the northern hemisphere summer over the 16 sites, are sufficiently consistent to propose using their all‐site, daily‐average values in a globally applicable soil‐moisture analysis scheme. Tests of this method at the First ISLSCP (International Satellite Land Surface Climatology Program) Field Experiment (FIFE) site confirm that soil‐moisture nudging can provide good estimates of near‐surface weather variables and surface fluxes in a numerical weather prediction (NWP) model which gives poor simulation of precipitation (thereby poor soil moisture). However, the PCA method is unable to give an accurate determination of the location of the soil moisture within soil layers that are accessible to the atmosphere. Further, these tests show that, if the NWP model has good simulation of precipitation but poor simulation of surface radiation, soil‐moisture nudging could wrongly vary the soil moisture so as to provide good estimates of the surface sensible‐heat flux, but poor estimates of surface latent‐heat flux. an approach to distinguish error sources from surface radiation and soil moisture is necessary for further improvement.

Accuracy of Wood Resistance Sensors for Measurement of Humidity

Abstract This work was undertaken to evaluate the accuracy of wood resistance sensors for measurement of relative humidity and to identify sources of error in this use. Relative humidity can be expressed as a function of the logarithm of the sensor's electrical resistance and of its temperature. We found that single-point calibration of each sensor compensates for most between-sensor variation, although care must be exercised during calibration. With careful calibration readings, error in relative humidity readings made with these sensors can be limited to ± 10% relative humidity under most conditions. The literature indicates that a lower degree of error than this is anticipated when electrical resistance is used to estimate moisture content measurements. Our data suggest that sorption hysteresis and sensor memory are significant contributors to this (± 10%) relative humidity error.

Passive microwave relative humidity retrievals using feedforward neural networks

A technique for retrieving atmospheric humidity profiles using passive microwave spectral observations from satellite and multilayer feedforward neural networks (MFNN) is introduced. Relative humidity retrievals on a global scale from simulated radiances at fifteen frequencies between 23.8 and 183.3 GHz yielded rms errors in relative humidity of 6-14% over ocean and 6-15% over land at pressure levels ranging from 131 mbar to 1013 mbar. Comparison with a combined statistical and physical iterative retrieval scheme shows that superior retrievals can be obtained at a lower computational cost using MFNN.

Comparative Experiments with MAPS on Different Parameterization Schemes for Surface Moisture Flux and Boundary-Layer Processes

This study compares several formulations parameterizing the surface moisture flux and boundary-layer processes using the θ-σ hybrid-b model of the Mesoscale Analysis and Prediction System (MAPS) within both 1D and 3D frameworks. A modified formula for computing the surface moisture flux is proposed based on the assumption that the layer below the lowest model computational level can be represented by three “physical” layers, of which the bottom one is the molecular layer. This three-layer aerodynamic (3LAD) scheme is compared with two-layer aerodynamic (2LAD) as well as flux matching and Penman-Monteith potential evapotranspiration (PM) schemes. Both a 10-day forecast period (3D) and case simulations demonstrate that the 3LAD scheme gives the best prediction in latent heat flux from the ground and mixing ratio in the atmosphere. The moisture flux produced by the 2LAD scheme is too large, especially over warm and moist surfaces. The mean 12-h forecast rms errors in relative humidity at the surface (10 m AGL) are 15.6%, 21.5%, and 26.0%, respectively, for the 3LAD, PM, and 2LAD schemes in a 10-day parallel test period using MAPS. For the boundary-layer parameterization, the Mellor-Yamada level 2.0 turbulence scheme (MY) and Blackadar convective scheme are compared. Results show that the MY scheme gives more reasonable boundary-layer structure and smaller rms forecast errors.