Search Results
You are looking at 1 - 10 of 48 items for
- Author or Editor: J. Jin x
- Refine by Access: All Content x
Abstract
The 55-yr California Cooperative Oceanic Fisheries Investigations (CalCOFI) dataset in the southern California Current reveals a significant surface-intensified warming and stratification (buoyancy frequency) change across the 1976/77 climate regime shift. However, the average depth of the thermocline, defined as the maximum gradient of temperature, did not change significantly across the regime shift. The maximum-gradient criterion for thermocline depth may be more appropriate than following an isotherm because the isotherm necessarily deepens in the presence of surface-intensified warming. As the surface heating changed the strength of stratification, it also changed the slope of the nitrate–temperature relation for the middepth waters (roughly 30–200 m). Thus, the quality of upwelled water may have been fundamentally altered after the shift.
Abstract
The 55-yr California Cooperative Oceanic Fisheries Investigations (CalCOFI) dataset in the southern California Current reveals a significant surface-intensified warming and stratification (buoyancy frequency) change across the 1976/77 climate regime shift. However, the average depth of the thermocline, defined as the maximum gradient of temperature, did not change significantly across the regime shift. The maximum-gradient criterion for thermocline depth may be more appropriate than following an isotherm because the isotherm necessarily deepens in the presence of surface-intensified warming. As the surface heating changed the strength of stratification, it also changed the slope of the nitrate–temperature relation for the middepth waters (roughly 30–200 m). Thus, the quality of upwelled water may have been fundamentally altered after the shift.
Abstract
Present-day OGCMs give low values of annual mean net heat flux (AMNHF) in the tropical Indian Ocean, compared to climatologies. AMNHF generation is examined in an open-boundary model of this region with realistic coastlines. In the first two of three experiments only annual mean wind stresses were applied so that a modified form of the “minimum depth criterion” of the previous paper would be applicable. Area-integrated AMNHF was well below observed values, despite the fact that western boundary inflow was substantially deeper and colder than was expected from the modified minimum depth estimate. The model showed large “spikes” in the gradient of “depth-integrated steric height” (DISH) along the western boundary, coinciding with coastline steps (which were absent in the previous paper). Most diapycnal entrainment occurred next to the coast, near these steps. In a third experiment a seasonal cycle of wind stress was added to the same annual mean. Annual mean diapycnal mixing and entrainment increased and the western boundary inflow deepened, resulting in substantially greater AMNHF for the same annual mean Ekman transports. However, area-integrated AMNHF was still well below the mean of directly observed surface fluxes. The recirculation around the “Great Whirl” doubled, permitting more cold water crossing the equator in one year to mix with recirculated water generated in a previous year. Entrainment up to the surface thus went by stages, over more than one year. The increased Great Whirl was related to stronger annual mean curls of nonlinear terms in the momentum equation, while the deeper entrainment was caused by stronger annual mean diapycnal mixing. In all experiments, diapycnal mixing was primarily due to the “flux corrected transport” (FCT) advective scheme, which in effect replaces spurious convective overturn by numerical diffusion. More research is needed to solve such problems, but sensitivity of AMNHF in OGCMs to time-varying forcing—due to seasonal, intraseasonal, or baroclinic instability—may offer a new source of climate predictability.
Abstract
Present-day OGCMs give low values of annual mean net heat flux (AMNHF) in the tropical Indian Ocean, compared to climatologies. AMNHF generation is examined in an open-boundary model of this region with realistic coastlines. In the first two of three experiments only annual mean wind stresses were applied so that a modified form of the “minimum depth criterion” of the previous paper would be applicable. Area-integrated AMNHF was well below observed values, despite the fact that western boundary inflow was substantially deeper and colder than was expected from the modified minimum depth estimate. The model showed large “spikes” in the gradient of “depth-integrated steric height” (DISH) along the western boundary, coinciding with coastline steps (which were absent in the previous paper). Most diapycnal entrainment occurred next to the coast, near these steps. In a third experiment a seasonal cycle of wind stress was added to the same annual mean. Annual mean diapycnal mixing and entrainment increased and the western boundary inflow deepened, resulting in substantially greater AMNHF for the same annual mean Ekman transports. However, area-integrated AMNHF was still well below the mean of directly observed surface fluxes. The recirculation around the “Great Whirl” doubled, permitting more cold water crossing the equator in one year to mix with recirculated water generated in a previous year. Entrainment up to the surface thus went by stages, over more than one year. The increased Great Whirl was related to stronger annual mean curls of nonlinear terms in the momentum equation, while the deeper entrainment was caused by stronger annual mean diapycnal mixing. In all experiments, diapycnal mixing was primarily due to the “flux corrected transport” (FCT) advective scheme, which in effect replaces spurious convective overturn by numerical diffusion. More research is needed to solve such problems, but sensitivity of AMNHF in OGCMs to time-varying forcing—due to seasonal, intraseasonal, or baroclinic instability—may offer a new source of climate predictability.
Abstract
A three-dimensional computational fluid dynamics (CFD) model is developed to simulate urban flow and dispersion, to understand fluid dynamical processes therein, and to provide practical solutions to some emerging problems of urban air pollution. The governing equations are the Reynolds-averaged equations of momentum, mass continuity, heat, and other scalar (here, passive pollutant) under the Boussinesq approximation. The Reynolds stresses and turbulent fluxes are parameterized using the eddy diffusivity approach. The turbulent diffusivities of momentum, heat, and pollutant concentration are calculated using the prognostic equations of turbulent kinetic energy and its dissipation rate. The set of governing equations is solved numerically on a staggered, nonuniform grid system using a finite-volume method with the semi-implicit method for pressure-linked equation (SIMPLE) algorithm. The CFD model is tested for three different building configurations: infinitely long canyon, long canyon of finite length, and orthogonally intersecting canyons. In each case, the CFD model is shown to simulate urban street-canyon flow and pollutant dispersion well.
Abstract
A three-dimensional computational fluid dynamics (CFD) model is developed to simulate urban flow and dispersion, to understand fluid dynamical processes therein, and to provide practical solutions to some emerging problems of urban air pollution. The governing equations are the Reynolds-averaged equations of momentum, mass continuity, heat, and other scalar (here, passive pollutant) under the Boussinesq approximation. The Reynolds stresses and turbulent fluxes are parameterized using the eddy diffusivity approach. The turbulent diffusivities of momentum, heat, and pollutant concentration are calculated using the prognostic equations of turbulent kinetic energy and its dissipation rate. The set of governing equations is solved numerically on a staggered, nonuniform grid system using a finite-volume method with the semi-implicit method for pressure-linked equation (SIMPLE) algorithm. The CFD model is tested for three different building configurations: infinitely long canyon, long canyon of finite length, and orthogonally intersecting canyons. In each case, the CFD model is shown to simulate urban street-canyon flow and pollutant dispersion well.
Abstract
Water vapor plays a key role in the global hydrologic cycle and in climatic change. However, the distribution and variability of water vapor in the troposphere are not understood well—in particular, in China with the complex Tibetan Plateau and the influence of the Asian and Pacific monsoons. In this paper, continuous global positioning system (GPS) observations for 2004–07 in China are used to produce precipitable water vapor (PWV) measurements; these measurements constitute the first investigation of PWV distribution and variability over China. It has been found that the stronger water vapor values are in southeastern China and the lower water vapor values are in northwestern China. These distributions are mainly affected by the latitude, topographical features, the season, and the monsoon. Water vapor variations over China are mainly dominated by seasonal variations. The strong seasonal cycles are in summer with maximum water vapor and in winter with minimum water vapor. The PWV in southeastern China has an annual amplitude of about 15 mm, much larger than in northwestern China at about 4 mm, and meanwhile the time of peak water vapor content is one month earlier than in other regions, probably because of the known rainy season (mei-yu). In addition, significant diurnal variations of water vapor are found over all GPS stations, with a mean amplitude of about 0.7 mm, and the peak value occurs around noon or midnight, depending on geographic location and topographical features. The semidiurnal cycle is weaker, with a mean amplitude of about 0.3 mm, and the first peak PWV value appears around noon.
Abstract
Water vapor plays a key role in the global hydrologic cycle and in climatic change. However, the distribution and variability of water vapor in the troposphere are not understood well—in particular, in China with the complex Tibetan Plateau and the influence of the Asian and Pacific monsoons. In this paper, continuous global positioning system (GPS) observations for 2004–07 in China are used to produce precipitable water vapor (PWV) measurements; these measurements constitute the first investigation of PWV distribution and variability over China. It has been found that the stronger water vapor values are in southeastern China and the lower water vapor values are in northwestern China. These distributions are mainly affected by the latitude, topographical features, the season, and the monsoon. Water vapor variations over China are mainly dominated by seasonal variations. The strong seasonal cycles are in summer with maximum water vapor and in winter with minimum water vapor. The PWV in southeastern China has an annual amplitude of about 15 mm, much larger than in northwestern China at about 4 mm, and meanwhile the time of peak water vapor content is one month earlier than in other regions, probably because of the known rainy season (mei-yu). In addition, significant diurnal variations of water vapor are found over all GPS stations, with a mean amplitude of about 0.7 mm, and the peak value occurs around noon or midnight, depending on geographic location and topographical features. The semidiurnal cycle is weaker, with a mean amplitude of about 0.3 mm, and the first peak PWV value appears around noon.
Abstract
The global response to tropical heating is studied by performing a time integration of a 15-level primitive equation model, starting with a basic flow maintained by a constant forcing. The direct, quasi-steady response to the tropical heating is seen during the first 20 days before baroclinic instability dominates. This technique enables the investigation of a variety of basic flows, from a resting state to a December–February 3D time-mean flow; of the timescales for establishing remote responses; and of nonlinear effects. It also allows the determination of timescales for the establishment of the response. The Gill-type response is seen in the lower troposphere in all cases. In the upper troposphere, depending on the basic conditions, the simple tropical quadrupole response of the Gill model shows considerable modification. The anticyclonic pair can be centered over the heating and can vary substantially in magnitude and vertical extent. The Rossby wave source and the upper-tropospheric divergence above the beating region is always found, but the existence and relative magnitudes of local Hadley and Walker cells as measured by upper-tropospheric convergence are strong functions of the flow. Both the Rossby wave source and the Rossby wave propagation are also strongly influenced by the ambient flow. Wave patterns extend to the equator in the regions in which the basic westerlies extend to the equator. Significant tropical zonal flow variations, which are also very dependent on the basic flow and the position of the heating, are also produced. The tropical and midlatitude response are generally established within a week. In an additional week the high-latitude pattern is determined and the subtropical wave pattern propagates back into the Tropics in the westerly wind regions. Nonlinear effects are found to be minor in all cases before the middle-latitude transients develop. On the two-week timescale of interest here, the sensitivity of steady-state models to the dissipations employed and to the existence of low-frequency modes is not found.
Abstract
The global response to tropical heating is studied by performing a time integration of a 15-level primitive equation model, starting with a basic flow maintained by a constant forcing. The direct, quasi-steady response to the tropical heating is seen during the first 20 days before baroclinic instability dominates. This technique enables the investigation of a variety of basic flows, from a resting state to a December–February 3D time-mean flow; of the timescales for establishing remote responses; and of nonlinear effects. It also allows the determination of timescales for the establishment of the response. The Gill-type response is seen in the lower troposphere in all cases. In the upper troposphere, depending on the basic conditions, the simple tropical quadrupole response of the Gill model shows considerable modification. The anticyclonic pair can be centered over the heating and can vary substantially in magnitude and vertical extent. The Rossby wave source and the upper-tropospheric divergence above the beating region is always found, but the existence and relative magnitudes of local Hadley and Walker cells as measured by upper-tropospheric convergence are strong functions of the flow. Both the Rossby wave source and the Rossby wave propagation are also strongly influenced by the ambient flow. Wave patterns extend to the equator in the regions in which the basic westerlies extend to the equator. Significant tropical zonal flow variations, which are also very dependent on the basic flow and the position of the heating, are also produced. The tropical and midlatitude response are generally established within a week. In an additional week the high-latitude pattern is determined and the subtropical wave pattern propagates back into the Tropics in the westerly wind regions. Nonlinear effects are found to be minor in all cases before the middle-latitude transients develop. On the two-week timescale of interest here, the sensitivity of steady-state models to the dissipations employed and to the existence of low-frequency modes is not found.
Abstract
Coupled ocean-atmosphere models exhibit a variety of forms of tropical interannual variability that may be understood as different flow regimes of the coupled system. The parameter dependence of the primary bifurcation is examined in a “stripped-down” version of the Zebiak and Cane model using the equatorial band approximation for the sea surface temperature (SST) equation as by Neelin. In Part I of this three-part series, numerical results are obtained for a conventional semispectral version; Parts II and III use an integral formulation to generate analytical results in simplifying limits. In the uncoupled case and in the fast-wave limit (where oceanic adjustment occurs fast compared to SST time scales), distinct sets of modes occur that are primarily related to the time scales of SST change (SST modes) and of oceanic adjustment (ocean-dynamics modes). Elsewhere in the parameter space, the leading modes are best characterized as mixed SST/ocean-dynamics modes; in particular, the continuous surfaces in parameter space formed by the eigenvalues of each type of mode can join.
A regime in the fast-wave limit in which the most unstable mode is purely growing, with SST anomalies in the eastern Pacific, proves to be a useful starting point for describing these mergers. This mode is linked to several oscillatory regimes by surfaces of degeneracy in the parameter space, at which two degrees of freedom merge. Within the fast-wave limit, changes in parameters controlling the strength of the surface layer or the atmospheric structure produce continuous transition of the stationary mode to propagating modes. Away from the fast-wave limit, the stationary mode persists at strong coupling even when time scales of ocean dynamics become important. On the weaker coupling side, the stationary mode joins to an oscillatory mode with mixed properties, with a standing oscillation in SST whose growth and spatial form may be understood from the SST mode at the fast-wave limit but whose period depends on subsurface oceanic dynamics. The oceanic dynamics, however, is only remotely related to that of the uncoupled problem. In fact, this standing-oscillatory mixed mode is insensitive to low-coupling complications involving connections to a sequence of uncoupled ocean modes at different parameter values, most of which are members of a discretized scattering spectrum. The implication that realistic coupled regimes are best understood from strong rather than weak coupling is pursued in Parts II and III. The interpretation of the standing-oscillatory regime as a stationary SST mode perturbed by wave dynamics gives a rigorous basis to the original physical interpretation of a simple model of Suarez and Schopf. However, viewing the connected modes as different regimes of a mixed SST/ocean-dynamics mode allows other simple models to be interpreted as alternate approximations to the same eigensurface; it also makes clear why varying degrees of propagating and standing oscillation can coexist in the same coupled mode.
Abstract
Coupled ocean-atmosphere models exhibit a variety of forms of tropical interannual variability that may be understood as different flow regimes of the coupled system. The parameter dependence of the primary bifurcation is examined in a “stripped-down” version of the Zebiak and Cane model using the equatorial band approximation for the sea surface temperature (SST) equation as by Neelin. In Part I of this three-part series, numerical results are obtained for a conventional semispectral version; Parts II and III use an integral formulation to generate analytical results in simplifying limits. In the uncoupled case and in the fast-wave limit (where oceanic adjustment occurs fast compared to SST time scales), distinct sets of modes occur that are primarily related to the time scales of SST change (SST modes) and of oceanic adjustment (ocean-dynamics modes). Elsewhere in the parameter space, the leading modes are best characterized as mixed SST/ocean-dynamics modes; in particular, the continuous surfaces in parameter space formed by the eigenvalues of each type of mode can join.
A regime in the fast-wave limit in which the most unstable mode is purely growing, with SST anomalies in the eastern Pacific, proves to be a useful starting point for describing these mergers. This mode is linked to several oscillatory regimes by surfaces of degeneracy in the parameter space, at which two degrees of freedom merge. Within the fast-wave limit, changes in parameters controlling the strength of the surface layer or the atmospheric structure produce continuous transition of the stationary mode to propagating modes. Away from the fast-wave limit, the stationary mode persists at strong coupling even when time scales of ocean dynamics become important. On the weaker coupling side, the stationary mode joins to an oscillatory mode with mixed properties, with a standing oscillation in SST whose growth and spatial form may be understood from the SST mode at the fast-wave limit but whose period depends on subsurface oceanic dynamics. The oceanic dynamics, however, is only remotely related to that of the uncoupled problem. In fact, this standing-oscillatory mixed mode is insensitive to low-coupling complications involving connections to a sequence of uncoupled ocean modes at different parameter values, most of which are members of a discretized scattering spectrum. The implication that realistic coupled regimes are best understood from strong rather than weak coupling is pursued in Parts II and III. The interpretation of the standing-oscillatory regime as a stationary SST mode perturbed by wave dynamics gives a rigorous basis to the original physical interpretation of a simple model of Suarez and Schopf. However, viewing the connected modes as different regimes of a mixed SST/ocean-dynamics mode allows other simple models to be interpreted as alternate approximations to the same eigensurface; it also makes clear why varying degrees of propagating and standing oscillation can coexist in the same coupled mode.
Abstract
The properties of the eigenmodes of the coupled tropical ocean-atmosphere system, linearized about a climatological basic state—and hence of the first bifurcation, which strongly determines the nature of the interannual variability, such as El Niño—show considerable dependence on the parameters of the coupled system. These eigenmodes are examined in a modified shallow-water model with simplified mixed-layer dynamics and a sea surface temperature (SST) equation, coupled to a simple atmospheric model. The model is designed so as to make analytical approximations feasible in various limits, as in a previous study by Neelin where the x-periodic case was analyzed. The realistic case of a finite ocean basin is treated here. An integral formulation of the eigenvalue problem is derived that provides a basis for making consistent approximations that include the effects of atmospheric and oceanic boundary conditions. We provide a scaling analysis to select parameters that give the most succinct insights into the behavior of the system, and outline the portions of this parameter space that are accessible to analytic results through the limits explored here and in Part III of this study. Important limits include the fast-wave limit, the limit where the time scale of ocean adjustment is fast compared to the time scale of SST change by coupled processes, and its converse, the fast-SST limit. The region of validity of the weak-coupling limit overlaps both of these, while that of the strong-coupling limit overlaps the fast-SST limit and approaches the region of validity of the fast-wave limit without a formal matching region.
In this part, we examine the weak-coupling limit, in which one expects the modes to be most closely related to those of the uncoupled problem. Here we treat two classes of mode from the uncoupled case: the SST modes (related to the time derivative of the SST equation) and the discrete modes from the ocean-dynamics spectrum, the ocean basin modes. From the numerical results of Part I, we know that away from the weak-coupling and fast-wave limits, the continuous surfaces in parameter space formed by the eigenvalues of each type of mode are joined, so that through most of parameter space the coupled modes are best characterized as mixed SST/ocean-dynamics modes. Series solutions for the weakly coupled modes are found to have radii of convergence that extend over modest but significant ranges of coupling values. The transition from the uncoupled modes to the fundamentally coupled mixed modes is examined. For the SST modes, coupling effects come to dominate the structure of basin-scale modes even at tiny coupling values. The structure of the ocean basin modes persists over a perceptible range of coupling, but structure changes involving the SST equation enter importantly as coupling is increased and the transition to mixed-mode structure occurs at small coupling, well within the range of the weak-coupling limit. This suggests that intuition and terminology borrowed from the uncoupled system is of limited value in analyzing coupled models and that it is more productive to consider prototype modes in fully coupled regimes.
Abstract
The properties of the eigenmodes of the coupled tropical ocean-atmosphere system, linearized about a climatological basic state—and hence of the first bifurcation, which strongly determines the nature of the interannual variability, such as El Niño—show considerable dependence on the parameters of the coupled system. These eigenmodes are examined in a modified shallow-water model with simplified mixed-layer dynamics and a sea surface temperature (SST) equation, coupled to a simple atmospheric model. The model is designed so as to make analytical approximations feasible in various limits, as in a previous study by Neelin where the x-periodic case was analyzed. The realistic case of a finite ocean basin is treated here. An integral formulation of the eigenvalue problem is derived that provides a basis for making consistent approximations that include the effects of atmospheric and oceanic boundary conditions. We provide a scaling analysis to select parameters that give the most succinct insights into the behavior of the system, and outline the portions of this parameter space that are accessible to analytic results through the limits explored here and in Part III of this study. Important limits include the fast-wave limit, the limit where the time scale of ocean adjustment is fast compared to the time scale of SST change by coupled processes, and its converse, the fast-SST limit. The region of validity of the weak-coupling limit overlaps both of these, while that of the strong-coupling limit overlaps the fast-SST limit and approaches the region of validity of the fast-wave limit without a formal matching region.
In this part, we examine the weak-coupling limit, in which one expects the modes to be most closely related to those of the uncoupled problem. Here we treat two classes of mode from the uncoupled case: the SST modes (related to the time derivative of the SST equation) and the discrete modes from the ocean-dynamics spectrum, the ocean basin modes. From the numerical results of Part I, we know that away from the weak-coupling and fast-wave limits, the continuous surfaces in parameter space formed by the eigenvalues of each type of mode are joined, so that through most of parameter space the coupled modes are best characterized as mixed SST/ocean-dynamics modes. Series solutions for the weakly coupled modes are found to have radii of convergence that extend over modest but significant ranges of coupling values. The transition from the uncoupled modes to the fundamentally coupled mixed modes is examined. For the SST modes, coupling effects come to dominate the structure of basin-scale modes even at tiny coupling values. The structure of the ocean basin modes persists over a perceptible range of coupling, but structure changes involving the SST equation enter importantly as coupling is increased and the transition to mixed-mode structure occurs at small coupling, well within the range of the weak-coupling limit. This suggests that intuition and terminology borrowed from the uncoupled system is of limited value in analyzing coupled models and that it is more productive to consider prototype modes in fully coupled regimes.
Abstract
In this paper, we investigate the relationship between upper ocean heat content (OHC) and El Niño–Southern Oscillation (ENSO) sea surface temperature (SST) anomalies mainly using the neutral recharge oscillator (NRO) model both analytically and numerically. Previous studies showed that spring OHC, which leads SST by 6–12 months, represents a major source of predictability for ENSO. It is suggested that this seasonality is caused by the seasonally varying growth rate in SST anomalies. Moreover, a shortened ENSO period will lead to a reduced SST predictability from OHC, with the most significant decrease occurring in the latter half of the calendar year. The cross-correlation relationship between OHC and ENSO SST anomalies is further identified in the damped and self-excited version of the recharge oscillator model. Finally, we suggest that the seasonal growth rate of ENSO anomalies is the cause of the seasonality in the effectiveness of OHC as a predictor in ENSO forecasting, particularly as it relates to the boreal spring persistence barrier and associated spring predictability barrier. We also explain the shorter lead time between spring OHC and ENSO SST anomalies after the turn of the twenty-first century in terms of the apparent higher frequency of the ENSO period.
Abstract
In this paper, we investigate the relationship between upper ocean heat content (OHC) and El Niño–Southern Oscillation (ENSO) sea surface temperature (SST) anomalies mainly using the neutral recharge oscillator (NRO) model both analytically and numerically. Previous studies showed that spring OHC, which leads SST by 6–12 months, represents a major source of predictability for ENSO. It is suggested that this seasonality is caused by the seasonally varying growth rate in SST anomalies. Moreover, a shortened ENSO period will lead to a reduced SST predictability from OHC, with the most significant decrease occurring in the latter half of the calendar year. The cross-correlation relationship between OHC and ENSO SST anomalies is further identified in the damped and self-excited version of the recharge oscillator model. Finally, we suggest that the seasonal growth rate of ENSO anomalies is the cause of the seasonality in the effectiveness of OHC as a predictor in ENSO forecasting, particularly as it relates to the boreal spring persistence barrier and associated spring predictability barrier. We also explain the shorter lead time between spring OHC and ENSO SST anomalies after the turn of the twenty-first century in terms of the apparent higher frequency of the ENSO period.
Abstract
Four independently developed high-resolution precipitation products [HRPPs; the Tropical Rainfall Measuring Mission (TRMM) Multisatellite Precipitation Analysis (TMPA), the Climate Prediction Center Morphing Method (CMORPH), Precipitation Estimation from Remotely Sensed Information using Artificial Neural Networks (PERSIANN), and the National Research Laboratory (NRL) blended precipitation dataset (NRL-blended)], with a spatial resolution of 0.25° and a temporal resolution of 3 h, were compared with surface rain measurements for the four summer seasons (June, July, and August) from 2003 to 2006. Surface measurements are 1-min rain gauge data from the Automated Weather Station (AWS) network operated by the Korean Meteorological Administration (KMA) over South Korea, which consists of about 520 sites. The summer mean rainfall and diurnal cycles of TMPA are comparable to those of the AWS, but with larger magnitudes. The closer agreement of TMPA with surface observations is due to the adjustment of the real-time version of TMPA products to monthly gauge measurements. However, the adjustment seems to result in significant overestimates for light or moderate rain events and thus increased RMS error. In the other three products (CMORPH, PERSIANN, and NRL-blended), significant underestimates are evident in the summer mean distribution and in scatterplots for the grid-by-grid comparison. The magnitudes of the diurnal cycles of the three products appear to be much smaller than those suggested by AWS, although CMORPH shows nearly the same diurnal phase as in AWS. Such underestimates by three methods are likely due to the deficiency of the passive microwave (PMW)-based rainfall retrievals over the South Korean region. More accurate PMW measurements (in particular by the improved land algorithm) seem to be a prerequisite for better estimates of the rain rate by HRPP algorithms. This paper further demonstrates the capability of the Korean AWS network data for validating satellite-based rain products.
Abstract
Four independently developed high-resolution precipitation products [HRPPs; the Tropical Rainfall Measuring Mission (TRMM) Multisatellite Precipitation Analysis (TMPA), the Climate Prediction Center Morphing Method (CMORPH), Precipitation Estimation from Remotely Sensed Information using Artificial Neural Networks (PERSIANN), and the National Research Laboratory (NRL) blended precipitation dataset (NRL-blended)], with a spatial resolution of 0.25° and a temporal resolution of 3 h, were compared with surface rain measurements for the four summer seasons (June, July, and August) from 2003 to 2006. Surface measurements are 1-min rain gauge data from the Automated Weather Station (AWS) network operated by the Korean Meteorological Administration (KMA) over South Korea, which consists of about 520 sites. The summer mean rainfall and diurnal cycles of TMPA are comparable to those of the AWS, but with larger magnitudes. The closer agreement of TMPA with surface observations is due to the adjustment of the real-time version of TMPA products to monthly gauge measurements. However, the adjustment seems to result in significant overestimates for light or moderate rain events and thus increased RMS error. In the other three products (CMORPH, PERSIANN, and NRL-blended), significant underestimates are evident in the summer mean distribution and in scatterplots for the grid-by-grid comparison. The magnitudes of the diurnal cycles of the three products appear to be much smaller than those suggested by AWS, although CMORPH shows nearly the same diurnal phase as in AWS. Such underestimates by three methods are likely due to the deficiency of the passive microwave (PMW)-based rainfall retrievals over the South Korean region. More accurate PMW measurements (in particular by the improved land algorithm) seem to be a prerequisite for better estimates of the rain rate by HRPP algorithms. This paper further demonstrates the capability of the Korean AWS network data for validating satellite-based rain products.
