1. Introduction
Precipitation is an important component of the water cycle and plays a critical role in the energy balance of Earth (Trenberth et al. 2009; Stephens et al. 2012; L’Ecuyer et al. 2015) as latent heat flux is commonly inferred from precipitation estimates. Spaceborne observations have complemented in situ data to estimate global precipitation amount and distribution for more than four decades. This has had a profound impact on understanding the state of Earth’s climate (Trenberth et al. 2007; Rodell et al. 2015) and advancing fields such as hydrology, atmospheric sciences, agriculture, and health. Climate models have also been used for decades to understand and predict future changes, yet often show large discrepancies compared to observed precipitation (Stephens et al. 2010; Sillmann et al. 2013; Behrangi and Richardson 2018).
While spaceborne precipitation estimation is important for guiding and assessing models, one should also realize that current precipitation observations also face large uncertainties (e.g., Adler et al. 2012; Behrangi et al. 2012a; Arabzadeh et al. 2020), and developing techniques and instruments to improve precipitation estimates are growing fields. The errors tend to be large over ocean, where in situ data are lacking to calibrate or bias adjust the products, especially in high latitudes (Adler et al. 2012; Behrangi et al. 2012a). This is important because about 77% of the global precipitation amount occurs over the oceans. Due to the significant limitation of in situ observations over ocean, satellites remain the main source for observation-based precipitation estimates there.
Behrangi et al. (2014) developed a Merged CloudSat, Tropical Rainfall Measuring Mission (TRMM) Precipitation Radar (PR), and Advanced Microwave Scanning Radiometer for Earth Observing System (AMSR-E) precipitation product (MCTA) that considers the entire precipitation histogram from drizzle and snowfall (from CloudSat) to intense precipitation (from PR). Using MCTA they estimated that the global oceanic precipitation is underestimated by about 4% for the V2.2 Global Precipitation Climatology Project (GPCP; Adler et al. 2003; Huffman et al. 2009) and about 9% for the Climate Prediction Center Merged Analysis of Precipitation (CMAP; Xie and Arkin 1997). This analysis was consistent with Rodell et al. (2015) and Trenberth et al. (2009), suggesting that increasing GPCP precipitation by about 5% can balance the surface water and energy budget of Earth. However, this increase was much smaller than the ∼15% considered by Stephens et al. (2012) to bring the surface energy budget into balance. Besides investigating global precipitation amount, studies have shown that precipitation errors are zonally dependent (Adler et al. 2012; Behrangi et al. 2012a; Behrangi and Wen 2017) and are largest over the midlatitude oceans in the Southern Hemisphere (SH) where GPCP V2.2, CMAP, and MCTA showed discrepancies of about 50% or more in determining zonal mean precipitation rate (Behrangi et al. 2014).
Major advances in precipitation estimation are being achieved through the Global Precipitation Measurement (GPM) mission (Skofronick-Jackson et al. 2017) and its Core Observatory satellite. The GPM Core Observatory was launched in February 2014 and carried two important instruments for precipitation measurements: 1) the Dual-Frequency Precipitation Radar (DPR) that includes Ku and Ka (13.6 and 35.5 GHz) bands (KuPR and KaPR, respectively) and 2) the GPM Microwave Imager (GMI) that has 13 channels with frequencies ranging from 10 to 183 GHz (Draper et al. 2015). These instruments have geographical coverage of ∼65°S–65°N (covering about 91% of the globe), which is about 40% more than TRMM’s coverage (∼35°S–35°N; covering ∼51% of the globe). The nominally better sensitivity of the DPR (∼0.2 mm h−1) than TRMM PR (∼0.5 mm h−1) and four additional high-frequency channels on the GMI (i.e., 166H, 166V, 183 ± 3V, and 183 ± 5V GHz) added new capabilities to detect and quantify snowfall and light rainfall. As precipitation retrieval from the combination of DPR and GMI is used in the Goddard profiling algorithm (GPROF; Kummerow et al. 1996; Randel et al. 2020) database for precipitation retrieval from passive microwave (PMW) radiometers (GMI and GPM constellation of PMW sensors), a global improvement of precipitation is expected. The combination of new sensors and retrieval methods should change previous estimates of global precipitation amount and distribution reported earlier (e.g., Behrangi et al. 2012a, 2014; Adler et al. 2017). Quantifying these changes is certainly insightful to a wide range of studies and applications that require accurate precipitation estimates.
In the present work, we assessed the amount and spatiotemporal distribution of oceanic precipitation rate from the latest version (i.e., V07) retrievals from the GPM Core Observatory instruments and a few other PMW sensors contributing to the GPM constellation sensors (Huffman et al. 2020); GPCP (i.e., V3.2; Huffman et al. 2023); and the latest version (V07) of the Merged CloudSat, TRMM, and GPM climatological precipitation product (MCTG; Behrangi and Song 2020), which is an update of the MCTA using GPM data. We compare these products with their previous versions and highlight the changes through zonal plots, geographical maps, and tables to quantify how the latest retrievals affect our estimation of regional and global precipitation rates over oceans. The outcomes will show regional and global changes in estimates of precipitation amounts, areas of progress, remaining discrepancies among the products, and will potentially guide the development of the next versions of precipitation products from individual or combined sensors, such as the Climate Prediction Center (CPC) morphing method (CMORPH; Joyce and Xie 2011), GPCP (Huffman et al. 2023), the Integrated Multi-satellitE Retrievals for GPM (IMERG; Huffman et al. 2020), and the Precipitation Estimation from Remotely Sensed Information Using Artificial Neural Networks (PERSIANN; Hsu et al. 1997), among others.
2. Method and datasets
We compare the latest two versions of monthly precipitation products from representative sensor types (i.e., radar-only, radar–radiometer combined, PMW imager, and PMW sounder) that are both of use individually and in producing multisensor precipitation products. A combination of zonal plots, spatial maps, and plots conditioned on environmental variables are employed to assess changes in the latest two versions of each product and to cross-compare various products. The zonal and global mean precipitation rates (using appropriate areal weighting) are also computed and reported in tabulated form. Three other precipitation products are investigated: European Center for Medium-Range Weather Forecasts Reanalysis 5 (ERA5), GPCP, and MCTG. The period of study is from May 2014 to April 2017 (3 years). This was set because all versions of GPM products were available to us during this period. Table 1 provides a list of the products used in this study. A brief description of these products is provided in the following subsections.
List of the products used in this study.
a. PMW conical and cross-tracking scanners
The products studied here include PMW conical scanners: GMI on the GPM Core Observatory, AMSR2 on the Japan Aerospace Exploration Agency Global Change Observation Mission–Water (GCOM-W1) satellite, and the Special Sensor Microwave Imager/Sounder (SSMIS) on the Defense Meteorological Satellite Program (DMSP) F17, and two PMW cross-tracking sounders: Microwave Humidity Sounder (MHS) on NOAA-19 and the Advanced Technology Microwave Sounder (ATMS) on the Suomi National Polar-Orbiting Partnership (SNPP) spacecraft. The V07 PMW products are based on GPROF V07 (Kummerow 2022), also known as GPROF 2021, a Bayesian approach that uses the GPM Combined Radar–Radiometer Algorithm (CORRA) to create its a priori databases. The previous version is GPROF V05. In GPROF V07, the a priori databases were constructed from the GPM Combined Radar–Radiometer Algorithm (V07) where it shows precipitation. Because the combined product may not have enough sensitivity to capture light precipitation, drizzle and light precipitation (up to 0.2 mm h−1) are added to the GPROF database from the operational Microwave Integrated Retrieval System (MiRS) Optimal Estimation retrieval (Liu et al. 2020). This results in GPROF V07 producing more precipitation than the Combined V07 product. Furthermore, GPROF V07 uses ERA5 reanalysis precipitation data over sea ice and for determining sea ice and ocean boundary, which was not the case for V05 (Kummerow 2022).
b. GPM Dual-Frequency Precipitation Radar (DPR)
Starting with V07, the KaPR scan pattern changes (since May 2018) were implemented, making it possible to apply dual-frequency KaPR and KuPR to the entire observation swath. Some of the changes in V07 compared to V06 are 1) improvement in the sidelobe clutter removal routine for the single-frequency radar algorithms, 2) changes in the hydrometeor drop size distribution (DSD) solver module (e.g., the relationship between precipitation rate and volume-weighted mean drop size and revision of DSD database), and 3) addition of the soil moisture effect on retrieval, all of which led to increased precipitation amounts over land. More detailed information is provided in Iguchi and Meneghini (2021). Here, only KuPR was included in the analysis as it provides data for a wider swath (i.e., 250 km compared to 125 km from KaPR prior to the scan pattern change), covering the entire GPM and TRMM eras. Note that our analysis of the DPR precipitation rate over ocean suggests that KuPR has higher precipitation rate in high latitudes (e.g., poleward of 40°N/S) than KaPR and DPR, while DPR shows higher precipitation rate in lower latitudes (Fig. S1 in the online supplemental material). This is the case for both full- and narrow-swath products, although the full-swath product shows slightly lower average precipitation in most zones and across all DPR-based precipitation products (Fig. S1). Note that neither KuPR nor KaPR are directly used in GPROF V07 nor in the multisensor products such as IMERG. The DPR contribution to these products is through the GPM CORRA product discussed as follows.
c. GPM combined radar–radiometer products
In principle, the CORRA product should produce the most accurate precipitation estimate as it uses coincident measurements from GPM radiometer and radar. Therefore, CORRA is popular for applications such as intercalibration of radiometers in IMERG (Huffman et al. 2020) and for construction of the GPROF V07 database, among others. CORRA also uses other datasets such as elevation, surface type, environmental parameters, and atmospheric vapor density and cloud water content (Olson 2022). Some of the main changes from V06 to V07 include greater a priori constraint on precipitation size distributions, and adding a near-surface clutter-zone correction to estimate surface precipitation. Two CORRA products are available: CORRA Ku that uses Ku radar and radiometer (GMI or TMI) and CORRA KuKa that uses dual-frequency radar and radiometer information. CORRA Ku is applied to the entire TRMM and GPM era, but CORRA KuKa is limited to the GPM era (when Ka is available). Furthermore, CORRA Ku provides full-swath product (250 km) throughout the record, while CORRA KuKa can produce full-swath retrieval only after the Ka scan pattern changes in 2018. IMERG uses CORRA Ku to benefit from the consistent longer record and larger samples of the product for the dynamic intercalibration of the radiometer-based precipitation estimates. However, GPROF V07 uses CORRA KuKa in its database.
Figure S2 shows that the CORRA Ku and CORRA KuKa products produce very similar mean precipitation rates over the oceans for all zones. However, similar to KuPR, mean precipitation tends to be slightly lower for the full-swath than the narrow-swath product.
Note that the GES DISC website that hosts the products makes the monthly CORRA products available under the name GPM-3CMB. Therefore, hereafter we use CMB instead of CORRA to reduce potential confusion.
d. MCTG
The MCTG climatology uses a combination of light rainfall and snowfall estimates from CloudSat and moderate to intense precipitation from the combined KuPR–radiometer products available from TRMM and GPM (Behrangi and Song 2020). The merging process is performed at each 0.5° × 0.5° grid over ocean at annual and seasonal scales to provide precipitation climatology maps and zonal distributions. The merging is performed by constructing the climatology-based normalized histograms of precipitation volume as a function of precipitation rate, then determining the precipitation rate at which CloudSat shifts from providing higher contributions to lower than the combined CMB (Behrangi and Song 2020). Poleward of 65°S, N, where GPM products are not available, MCTG uses CloudSat rain and snowfall products alone. Analysis has shown that CloudSat’s 94-GHz (W band) radar almost never experiences signal saturation (Tanelli et al. 2008) for precipitation events that occur in this region (Behrangi et al. 2012a). However, CloudSat coverage is limited to 81°S/N, so MCTG is only produced within 81°S–81°N. The latest MCTG uses CMB Ku V07 and is referred to MCTG V07, versus MCTG V06 that used CMB Ku V06. The MCTG V07 product used here is constructed from the first three years of the GPM products from April 2014 to March 2017 that matches the study period.
e. GPCP
GPCP is a combined satellite–gauge precipitation dataset that emphasizes the long-term standards of consistency and homogeneity. GPCP uses the Global Precipitation Climatology Centre (GPCC) as its in situ component over land and satellite data over both land and ocean. It uses selected SSMI, SSMIS, geostationary infrared imagers, and polar-orbiting infrared sounders. Some of the major changes of the latest GPCP (V3.2) dataset compared to its previous version (V2.3) include higher spatial resolution (0.5° × 0.5° at daily and monthly scales, instead of 2.5° × 2.5° at monthly and 1.0° × 1.0° at daily scales), using GPROF for precipitation retrieval from SSMI and SSMIS, using MCTG climatology over the mid- and high-latitude oceans and an updated Tropical Composite Climatology (TCC; Adler et al. 2009; Wang et al. 2014) over the tropical oceans for climatological calibration of the GPCP, modifying the gauge undercatch correction scheme over Eurasia and North Asia based on snowfall accumulation insights from the Gravity Recovery and Climate Experiment (GRACE) mass change observations (Behrangi et al. 2018, 2019), and improving the overall consistency of the data record, especially the infrared sounders in high latitudes (Huffman et al. 2023).
f. ERA5
ERA5, released by the European Centre for Medium-Range Weather Forecasts (ECMWF), is the fifth generation of global climate reanalysis data and is the successor of ERA-Interim (Hersbach et al. 2020). ERA5 covers Earth on 30-km grids and resolves the atmosphere using 137 levels from the surface up to a height of 80 km. Like the previous ERA-Interim, it uses a 4D variational assimilation scheme. By employing the improved Earth system model and data assimilation method, ERA5 was improved over ERA-Interim (Albergel et al. 2018; Hoffmann et al. 2019). ERA5 has a better global balance of precipitation and evaporation than ERA-Interim. Here we use sea surface temperature (SST) and total precipitable water (TPW) from ERA5 for regime-dependent precipitation analysis.
3. Results
a. Zonal mean and geographical maps
Figure 1 shows the zonal distribution of annual mean precipitation rate over the oceans from the GPM Core Observatory instrument products (KuPR, GMI, and CMB) for the latest (V07) and previous versions. GMI and CMB V07 products have higher precipitation rate than the earlier versions (i.e., V05 and V06, respectively) across the entire zone (65°S–65°N) covered by the GPM Core Observatory. However, the V07 KuPR product shows lower precipitation rates than V06 at every latitude. Comparison of precipitation maps between KuPR V07 and V06 (Fig. 2b) shows that KuPR V07 decreases in most grids, but not everywhere. KuPR V07 shows precipitation increases near equator in the Pacific Ocean, higher latitudes in the Southern Ocean, Gulf of Alaska, and several other regions. In contrast, the geographical maps of mean precipitation rate of GMI and CMB products (Figs. 2c–f) show no reduction in precipitation rate at any grid over the oceans. The increase in precipitation ranges from near zero (mainly in the eastern side of subtropical high-pressure regions) to its largest values over the deep tropics and in high latitudes. Overall, for the 65°S–65°N zone, mean oceanic precipitation rate (V07) for KuPR, GMI, and CMB change by −8.07%, +12.76%, and +9.03% compared to their previous versions, respectively (Table 2). Note that while KuPR V07 is moving away from the other estimates over the oceans, the comparison over land with WSR-88D shows that KuPR V07 is more consistent with WSR-88D than V06, partly due to the systematic reduction in reflectivity values in the new version (Li et al. 2023).
Zonal distributions of average oceanic precipitation rate from GPM Core Observatory V07 and previous (V05 or V06) products. A zonal bin size of 0.25° is used.
Citation: Journal of Hydrometeorology 25, 2; 10.1175/JHM-D-23-0082.1
(left) Maps of average oceanic-precipitation rate from GPM Core Observatory V07 products and (right) comparison with their corresponding previous version for (a),(b) KuPR; (c),(d) GMI; and (e),(f) combined (CMB) products.
Citation: Journal of Hydrometeorology 25, 2; 10.1175/JHM-D-23-0082.1
Global and zonal mean oceanic precipitation rates from products examined in this study and comparison to their previous versions. GPM V07 products mean (μ) and standard deviation (σ) are shown in the bottom of the table. Because MCTG covers 81°S–81°N, MCTG averages and all comparisons with MCTG are limited to 81°S–81°N (shown as 81°S/N in the table). Note that the bold fonts are for the new (latest) version of the products.
Similarly, Fig. 3 shows the zonal distribution of oceanic precipitation rate for a selection of GPM PMW constellation members for the two most recent versions of the products. These instruments are on polar-orbiting satellites and cover 90°S–90°N. It can be seen that, on average, precipitation rate in V07 increased for almost all zones for all the PMW products, although for the PMW sounders (i.e., MHS and ATMS) there is a slight reduction in mean precipitation rate around 35°–45° latitude in both hemispheres. Compared to the other products, MHS and ATMS show much larger increases in peak precipitation intensity in the tropics around 5°N. Note that MHS and ATMS are cross-tracking sounders, while AMSR-2 and SSMIS are conical scanners (as is the GMI). By putting all the products in one plot, Fig. 3 shows that the type of sensor helps explain systematic differences in zonal averages. The differences are more obvious within the 30°–50° latitude zone in both hemispheres and near 5°N. The CMB product is used as a transfer agent in the figure for comparing the precipitation rates from the constellation sensors (Fig. 3) and the GPM Core Observatory instruments (Fig. 1). PMW products show a slight shift of precipitation peak near 40°S toward higher latitudes, compared to CMB and KuPR products. This is also the case for GMI compared to both CMB and KuPR (Fig. 1).
Zonal distributions of average oceanic precipitation rate from GPM constellation sensors V05 and V07 products, all in one plot together with CMB V07 for comparison with GPM Core Observatory plots shown in Fig. 1. A zonal bin size of 0.25° is used.
Citation: Journal of Hydrometeorology 25, 2; 10.1175/JHM-D-23-0082.1
The geographical maps of annual mean precipitation rates (Fig. 4) provide additional information related to regional differences. Similar to GMI, SSMIS and AMSR-2 show no decrease in precipitation rate for V07 compared to V05. However, MHS and ATMS V07 precipitation products show a reduction in precipitation rate in most mid latitude regions compared to V05. This reduction is often larger in ATMS than MHS. A closer look at the difference maps for all of the PMW products (including GMI; Fig. 2d) shows a narrow band of near-zero change in precipitation rate over ocean around 55°S–60°S, near the coast of Antarctica. This is the case for both PMW imagers and sounders. In consultation with the GPROF team (C. Kummerow and P. Brown 2023, personal communication) we conclude that this band around the sea ice edge is a result of the low precipitation rates over high-latitude ocean. Specifically, where that region is classified as ocean, the precipitation rates drop to almost zero around 55°–60°S due to the very low CMB precipitation rates that are used in the GPROF database in this region. However, the surface areas classified as sea ice have increased precipitation due to the use of ERA5 precipitation in V07 (Kummerow 2022). The shift between these surface-type precipitation rates creates the near-zero band of lower precipitation that is nearly white-colored in the difference plots.
(left) Maps of average oceanic precipitation rate from GPM constellation sensors V07 products and (right) comparison with their corresponding V05 (V07 minus V05) for (a),(b) SSMIS on DMSP F17; (c),(d) MHS on NOAA-19; (e),(f) ATMS on SNPP; and (g),(h) AMSR2 on GCOM-W1.
Citation: Journal of Hydrometeorology 25, 2; 10.1175/JHM-D-23-0082.1
It should be noted that part of the widespread increase in precipitation rate in PMW V07 compared to V05 is related to addition of light precipitation rates in the PMW V07 products from MiRS over ocean (section 2a). Overall, the global ocean’s (90°S–90°N) mean annual precipitation rates from the version 07 of the SSMIS, MHS, ATMS, and AMSR2 are 18.01%, 14.76%, 11.90%, and 12.6% higher than their corresponding V05 mean precipitation rates (Table 2). Table 2 also shows similar analysis for 65°S–65°N and 25°S–25°N ocean regions.
b. Comparison with MCTG
Comparison with in situ observations can provide some useful insights (e.g., Bolvin et al. 2021), but in situ observations are very limited over the oceans and often limited to lower latitudes. Alternatively, the products can be compared with MCTG V07 (Behrangi and Song 2020), providing gridded maps of annual mean precipitation rate over the oceans by including very low (i.e., drizzle and snowfall) to intense precipitation rates from advanced sensors. The annual map of MCTG V07 mean precipitation rate over the oceans and changes from MCTG V06 are shown in Fig. 5. The pattern of precipitation increase for MCTG (Fig. 5b) is similar to that of CMB (Fig. 2f), but with smaller magnitude, especially in high latitudes. This is because MCTG V06 already used the CloudSat product to compensate for drizzle, light rainfall, and snowfall that are frequent in high latitudes and were largely missed by CMB V06 (Behrangi and Song 2020). Therefore, CMB V07 minus CMB V06 (Fig. 2f) shows larger values than MCTG V07 minus MCTG V06 (Fig. 5b).
(a) Maps of mean oceanic precipitation rate from MCTG V07 and (b) MCTG V07 minus MCTG V06 at 0.5° × 0.5°. Note that MCTG covers global oceans up to 81°S/N.
Citation: Journal of Hydrometeorology 25, 2; 10.1175/JHM-D-23-0082.1
Figure 6 displays the difference between the mean annual precipitation map from MCTG V07 and precipitation rates from GPM Core Observatory instruments (Figs. 6a,c,e) and constellation sensors (Figs. 6b,d,f,g). Precipitation rate from MCTG V07 is equal or higher than CMB V07 as expected. Also, MCTG precipitation rate is equal to or higher than KuPR V07 (Fig. 1e), showing even larger differences than with CMB V07 (Fig. 1a). However, this is not necessarily the case for the PMW products. While MCTG V07 shows higher precipitation rates than PMW products in most places, precipitation rates from PMW sensors tend to be slightly larger than MCTG over the intense precipitation regions of the ITCZ (especially noticeable over the Pacific Ocean, centered at ∼5°N, and the northern part of the Indian Ocean). This difference is larger for the cross-tracking sounders (i.e., ATMS and MHS) than the conical scanners [i.e., AMSR-2, GMI (not shown), and SSMIS]. PMW sensors also show higher mean precipitation rates than MCTG in the Arctic regions, where sea ice is present. In this region MCTG precipitation rate comes from combination of rainfall and snowfall from CloudSat. The reason that PMW sensors, with lower sensitivity to light rain and snowfall than CloudSat, estimate higher precipitation rate than CloudSat could be partly related to the fact that the latest GPROF algorithm (used in V07) uses ERA5 precipitation rate in its database over sea ice (Kummerow 2022). In addition, it is very difficult to distinguish between snow and ice particles in clouds over frozen surfaces using PMW sensors, so high errors are expected. Nonetheless, CloudSat may miss shallow precipitation due to ground clutter (Tanelli et al. 2008), which can lead to underestimation of precipitation, especially in high latitudes (Behrangi et al. 2018).
Geographical comparison of mean annual precipitation rates from (a),(c),(e) GPM Core Observatory instruments and (b),(d),(f),(g) constellation of PMW sensors with MCTG V07.
Citation: Journal of Hydrometeorology 25, 2; 10.1175/JHM-D-23-0082.1
Figure 7 compares zonal distribution of mean precipitation rate from MCTG V06, MCTG V07, ERA5, the latest GPCP (V3.2), and the V07 satellite products over the oceans. Based on Fig. 7, 1) mean precipitation rate from MCTG V07 is only slightly higher than MCTG V06 and it very much follows the MCTG V06 pattern, suggesting that CloudSat (used in MCTG V06) provides part of the information that CMB V07 introduced, compared to CMB V06; 2) GPCP V3.2 closely follows MCTG V06 and that is reasonable as MCTG V06 is used to guide the month-of-the-year oceanic precipitation rate climatology in GPCP V3.2 (Huffman et al. 2023); 3) precipitation products are least consistent in the midlatitude oceans in the SH (i.e., the 35°–55°S zone), displaying the largest spread in mean precipitation rate and location of peak precipitation; 4) the conical scanners (i.e., AMSR2, SSMIS) and CMB V07 show the precipitation peak near 40°S, but PMW sounders (MHS and ATMS; the latter not shown due to the similarity of the two) show a precipitation peak near 50°S that seems to agree more with ERA5 than other satellite precipitation products; and 5) all the precipitation products seem to agree on the location of the precipitation peaks (e.g., at ∼7°S, ∼5°N, and ∼40°N) and lows (∼20°S and 25°N) in other zones (i.e., excluding the Southern Ocean), at least within no more than a 2° shift in latitude. The analysis, however, shows that there are still large differences among GPCP V3.2, ERA5, and the GPM precipitation products in quantification of zonal mean precipitation rates (see Fig. S3). Note that inclusion of ERA5 in the analysis is motivated by its popularity and its relatively high skill over high-latitude ocean and sea ice (Song et al. 2020, 2021).
Zonal distribution of mean ocean precipitation rate from MCTG V06, MCTG V07, ERA5, the latest GPCP (V3.2), and V07 satellite products. To enhance clarity, ATMS is not shown as it is very similar to MHS.
Citation: Journal of Hydrometeorology 25, 2; 10.1175/JHM-D-23-0082.1
In Table 2, the mean precipitation rates of the products are compared to their previous versions and MCTG V07, where possible. As MCTG covers 81°S–81°N, all analysis related to MCTG are within 81°S–81°N, instead of 90°S–90°N. Based on the V07 products that cover the entire global ocean, mean oceanic precipitation within 90°S–90°N ranges from 3.01 mm day−1 (from ATMS) to 3.28 mm day−1 (from ERA5). The latest GPCP V3.2 falls in the middle (3.15 mm day−1) and shows a 6.78% increase compared to its previous version (V2.3). Similar results are also seen for 81°S–81°N, as the geographical area covered by 81°–90°N/S is too small to substantially change the global mean precipitation rates. Within 81°S–81°N, the largest ocean-average value is from MCTG V07 (3.38 mm day−1). ERA5 offers the second highest value (3.30 mm day−1), which is only 2.37% less than MCTG V07. This is also fairly similar for 65°S–65°N, where MCTG V07 continues to provide the highest estimate (3.51 mm day−1), followed by ERA5 (∼2.85% lower). GPCP V3.2 is 6.55% lower than MCTG V07, and GPM V07 products show a larger range from KuPR to SSMIS, displaying lower values than MCTG V07 by 25.36% and 4.84%, respectively.
The product spread in the 25°S–25°N zone is large and ranges from 3.15 mm day−1 (for KuPR V07) to 4.09 mm day−1 (for MHS). The orders are slightly different in the tropics (25°S–25°N), where ERA5 (3.89 mm day−1) shows 2.64% higher mean precipitation rate than MCTG 7 (3.79 mm day−1). This is because in this region MCTG highly depends on the CMB V07 product (3.59 mm day−1), which has much lower values than ERA5. ATMS V07 provides the second highest estimate (3.96 mm day−1) and GPCP V3.2 (3.49 mm day−1) falls in the middle of the range.
Table 2 shows that V07 products each have higher mean precipitation rates than the previous version, except KuPR. The highest increase in 90°S–90°N, 81°S–81°N, and 65°S–65°N regions is for SSMIS (18.01%, 17.88%, and 15.97%, respectively). In the 25°S–25°N zone, the highest increase is for MHS (19.24%) followed by ATMS (18.92%). MHS V07 and ATMS V07 also show the highest mean precipitation rate among all other products. The large increase of the sounders’ precipitation rate in 25°S–25°N needs further investigation, as other sensors do not show more than 10.48% increase and the increase is about 8% for GMI and AMSR2. The large decrease in KuPR V07 precipitation rate (8.07% in 65°S–65°N and 9.48 in 25°S–25°N compared to KuPR V06) also needs investigation, as it takes KuPR further from the other estimates in V07. By calculating mean (μ) and standard deviation (σ) of the GPM V07 products, it can be seen that GPCP V3.2 falls within μ ± 1σ in 90°S–90°N, 81°S–81°N, 65°S–65°N, and 25°S–25°N. ERA5 has higher values than GPCP V3.2 and exceeds μ + 2σ in 90°S–90°N and 81°S–81°N, exceeds μ + 1σ in 65°S–65°N, but falls within μ ± 1σ in 25°S–25°N (Table 2).
Similar quantitative analysis can be performed for different zones in the Northern Hemisphere (NH) and SH (Table S3). It can be seen that the increase in precipitation rate (from V05 to V07) in high latitudes is substantial, ranging from 69% (for AMSR2) to 126% (for SSMIS) at 65°–81°N and from 119% (for MHS) to 229% (for SSMIS) at 81°–65°S. While the V07 PMW products are still lower than MCTG V07 at 81°–65°S, they exceed MCTG V07 at 65°–81°N by up to ∼13% for SSMIS. In both high-latitude regions in SH and NH, ERA5 shows higher precipitation rate than MCTG 07 by ∼14% and 17%, respectively. Analysis of GPM V07 μ and σ shows that GPCP V3.2 and ERA5 are higher than μ by more than 2σ in 65°–25°S, while in 25°–65°N ERA5 and GPCP V3.2 are within μ ± 2σ. GPCP V3.2 exceeds μ + 2σ in 81°–65°S, but falls inside μ ± 1σ for 65°–81°N. ERA5 exceeds μ + 2σ in both 81°–65°S and 65°–81°N.
c. Annual cycle analysis
Figure 8 shows monthly mean precipitation variation (MPV), separately in SH and NH for middle (25°–65°) and high (65°–90°) latitudes. The old and new versions of the products are shown in dashed and solid lines, respectively. GPCP V3.2 and ERA5 are shown in bold solid lines. For the midlatitude region, clear and opposite seasonal cycles are observed in both NH (Fig. 8a) and SH (Fig. 8b). The products are generally consistent in capturing the timing of the monthly precipitation minimum and maximum. However, this is not the case for the PMW sounder products (MHS and ATMS), regardless of their product version. In NH most products show the highest and lowest monthly precipitation rate in December–January and May–July, respectively. However, PMW sounders show the highest and lowest precipitation rates in October and April. Furthermore, their monthly precipitation rate is not as symmetric as other products in NH. Similarly in SH (Fig. 8b), MHS and ATMS show different monthly precipitation distribution compared to the other products with monthly maximum precipitation rate in March (as opposed to summertime in other products) and minimum precipitation rate in July (in contrast to December in most of the other products). Over 65°–25°S, the sensors relative amplitude (i.e., maximum minus minimum monthly precipitation rate divided by annual mean precipitation rate) range from 21.51% (for SSMIS V07) to 43.81% (for KuPR V07) (Table S1). These numbers are 34.02% and 31.91% for GPCP V3.2 and ERA5. For 25°–65°N, the sensors’ relative amplitudes range from 32.83% (for ATMS V07) to 63.84% (for GMI V07). These numbers are 60.00% and 52.47% for GPCP V3.2 and ERA5. In the colder months, vast areas in the NH high latitude are covered by sea ice. This suggests that compared to GPCP V3.2 and ERA5, the V07 products may underestimate precipitation rates over the sea ice in NH. Song et al. (2020) used airborne observation of snow depth over sea ice and showed that V05 products strongly underestimate snowfall accumulation, while ERA5 overestimates. With the major increase of precipitation rate in V07 (compared to V05), further studies are needed to assess the performance of the V07 products.
Monthly mean oceanic precipitation rates separately for mid- and high-latitude regions in the Northern and Southern Hemispheres.
Citation: Journal of Hydrometeorology 25, 2; 10.1175/JHM-D-23-0082.1
The largest MPV among the products occurs in the high latitudes (Figs. 8c,d for 65°–90°N and 90°–65°S, respectively). For 65°–90°N (Fig. 8c) the sensors’ relative amplitudes range from 90.48% (for SSMIS V07) to 106.61% (for MHS V07); these numbers are considerably smaller for GPCP V3.2 (67.86%) and ERA5 (65.51%). In this region, the PMW V07 products generally show a similar pattern with maximum and minimum precipitation rate occurring in boreal summer and winter, respectively. In contrast, GPCP V3.2 and ERA5 show a different monthly precipitation distribution with maximum and minimum precipitation occurring in September and April, respectively. GPCP V3.2 and ERA5 are generally consistent in capturing precipitation monthly distribution, although ERA5 tends to show a slightly higher monthly precipitation rate than GPCP in most months. The V07 products show a similar pattern to the previous version, but with a much larger precipitation rate than the previous versions. The V07 products even exceed both GPCP V3.2 and ERA5 in boreal spring and summer. Note that the relatively large spike of precipitation rate in AMSR2 V05 in July and August is not observed in V07.
In 90°–65°S (Fig. 8d) the sensors’ relative amplitudes range from 44.17% (for SSMIS V07) to 49.55% (for MHS V07); these numbers are 28.15% and 43.75% for GPCP V3.2 and ERA5, which are much smaller than their corresponding regions in NH (65°–90°N). ERA5 shows much larger precipitation rates than all other products and displays peak precipitation rate in April. GPCP V3.2 provides the second-highest precipitation rate in most months with peak precipitation rate in September. The V07 conical scanners (SSMIS and AMSR2) show similar monthly precipitation patterns with precipitation peaks in November, but SMMIS shows a higher precipitation rate than AMSR2. MHS and ATMS show almost identical monthly precipitation distribution with peak precipitation occurring in January. Besides the increase in magnitude, V07 products show a noticeable change in precipitation pattern compared to V05; especially for MHS and ATMS the maximum precipitation rate in March seen in V06 does not exist in V07.
Averaging together the opposing seasonal cycles in each hemisphere leads to a smoothed signal. MPV is relatively small when averaged over 90°S–90°N and 65°S–65°N (i.e., relative amplitude is less than 7%), but it is larger when averaged over 25°S–25°N (i.e., relative amplitude around ∼10%) as can be seen in Fig. S4 and Table S2. Again, KuPR V07 and CMB V07 show slightly different distributions for monthly precipitation rate compared to the other sensor products.
d. Comparison versus SST and TPW
Zonal, geographical, and annual comparisons are informative, but may not reveal the whole story due to the averaging over large regions and long periods. Regional and seasonal analysis could enhance the assessments, but that will be very lengthy as each region or season may have to be analyzed separately. An alternative approach is to condition the analysis on physical variables, which also represent regional and seasonal variations. Such analyses can also advance understanding of the mechanisms behind the observed differences and can be useful for diagnostic error analysis (e.g., Arabzadeh and Behrangi 2021). Here we use SST and TPW from ERA5 for comparing different precipitation products. SST has a strong connection with cloud types (Behrangi et al. 2012b) and at monthly scale is well correlated with near-surface air temperature (Feng et al. 2018). Note that SST is not a prognostic variable for ERA5, it is observed and is a prescribed boundary condition. TPW is also a good indicator for type of precipitation regime, as discussed in Berg et al. (2006), who found that differences in rainfall intensity between the TRMM Precipitation Radar (PR) and TRMM Microwave Imager (TMI) are highly correlated with TPW and the intensity–TPW relationship is fairly invariant over seasonal and interannual time scales. Note that GPROF, used here for PMW precipitation retrieval, also uses surface air temperature and TPW from ERA5 for precipitation regime classification over both land and ocean.
Figure 9 shows the results for precipitation rates binned by SST (Fig. 9a) and TPW (Fig. 9b) over the oceans within 65°S–65°N, where all of the products are available. Sample counts used in calculating average values in each bin are also shown in Figs. 9c and 9d for SST and TPW, respectively. For SSTs > 298 K, the precipitation rate versus SST relationship and the location of peak precipitation (at SST of about 303 K) are generally consistent with the earlier studies of the relationship between the fraction of deep convective clouds and SST in the tropics (Behrangi et al. 2012b; Waliser and Graham 1993), who demonstrated that for SSTs exceeding 302.5 K deep convection is reduced. Studies also suggest that within this range of SST the vertical stability of the tropical troposphere is sufficiently reduced to cause the onset of large-scale moist convection (Lau and Shen 1988; Betts and Ridgway 1989). It can be seen that precipitation products are generally less consistent (larger spread) in SST and TPW bins with low or high values. The majority of precipitation in the lower SSTs comes from stratocumulus and nimbostratus clouds, especially poleward of 50°N/S (Sassen et al. 2008; Behrangi et al. 2012b), while convective clouds produce a large fraction of precipitation at the highest SSTs. As discussed in Behrangi et al. (2012a), a large fraction of light precipitation rate from stratocumulus clouds may not even be detected by existing radar and PMW sensors. All V07 products have lower precipitation rates than both GPCP V3.2 and ERA5 at TPW < ∼10 mm or SST < ∼282 K (with the exception of GMI V07 at the lowest SST bin). Note that all of the V07 PMW precipitation products show a sharp increase at the lowest SST (i.e., near 270 K) bin that can be related to the addition of drizzle and light precipitation from MiRS product implemented in the V07 products (Kummerow 2022). This is not observed for KuPR and CMB products. In high latitudes, GPCP uses precipitation estimates from the Atmospheric Infrared Sounder (AIRS) on NASA’s Aqua satellite using the Susskind et al. (1997) retrieval technique, which seems to perform relatively well in high latitudes once the dataset is homogenized and calibrated (Bolvin et al. 2009; Behrangi et al. 2016, 2020). ERA5 tends to show higher precipitation rates than GPCP V3.2, except when ∼12 mm < TPW < ∼24 mm or 280 K < SST < 294 K.
Mean precipitation rate from various products over the ocean conditioned on (a) SST and (b) TPW and (c),(d) the corresponding numbers of samples used in the calculation. The bin sizes for SST and TPW plots are 2 K and 2 mm, respectively. The analysis is performed within 65°S–65°N to include all the studied products.
Citation: Journal of Hydrometeorology 25, 2; 10.1175/JHM-D-23-0082.1
All of the products agree on precipitation peak around SST of 302 K. However, for the second precipitation peak at lower SST, products differ. GPCP V3.2 and all the V07 products (except PMW sounders) show a precipitation peak at ∼286 K (Fig. 9a), but ERA5 sees the peak around 282 K. MHS and ATMS fall in the middle with precipitation peak at ∼284 K and display a different pattern than the other conical scanners and radar products, more like ERA5. The observed shift of ATMS, ERA5, and MHS precipitation peak to the lower SSTs (compared to the conical scanners and GPCP V3.2) is also seen in the TPW plot (Fig. 9b), in which MHS, ATMS, and ERA5 show the precipitation peak at ∼12 mm, while the other products show the peak at ∼15 mm. At lower TPWs and SSTs ATMS and MHS show smaller precipitation rates than other products (except KuPR V07 and CMB V07), at the very high TPWs and SSTs, they show higher mean rates than all of the other products. This is especially noticeable at the highest TPW bin around 60 mm, where the products show more than 30% spread in mean precipitation rate compared to a grand average (Fig. 9b). This is also the case for precipitation peak near 302 K (Fig. 9a). Nonetheless, the largest relative spread in the precipitation products (highest rate minus lowest rate divided by the mean) is at the lowest SSTs and TPWs, exceeding 100%, and pointing to an important area for future improvements. By analysis of standard deviations around mean precipitation rates, it can be seen that there are still large differences among GPCP V3.2, ERA5, and the GPM precipitation products in quantification of mean precipitation rates versus SST and TPW values (see Fig. S5).
The new and previous versions of precipitation products are compared in Fig. 10 (also see Fig. S6 for further clarity). A few points can be highlighted: CMB and conical scanners V07 products show a shift to higher precipitation rates than their previous versions while preserving a similar precipitation–SST (Fig. 10a) and precipitation–TPW (Fig. 10b) relationships (e.g., similar precipitation peak and distribution). This is not necessarily the case for the PMW sounders (only MHS is shown as ATMS is similar to MHS). For MHS, V07 shows lower precipitation rate for 282 K < SST < 294 K and 15 mm < TPW < 24 mm and much higher mean precipitation rates at SST < 282 K, SST near 302 K, and TPW > 55 mm. MHS V07 precipitation peak and pattern also deviates from MHS V05 and other precipitation products at lower SST and TPW bins. GMI V05 shows a significantly lower mean precipitation rate (than other products) at the highest TPW bin (∼60 mm), but this is not observed in GMI V07. Last, KuPR V07 tends to be lower than KuPR V06 at TPW > 24 mm and SST > 292 K and higher than V06 at lower TPWs and SSTs, where stratocumulus clouds are most frequent. This may provide some insights into the reason why KuPR V07 has an overall lower mean precipitation rate than KuPR V05.
As in Fig. 9, but comparing current and previous versions of GPM and GPCP products.
Citation: Journal of Hydrometeorology 25, 2; 10.1175/JHM-D-23-0082.1
The binning of all rainfall from 65°S to 65°N together and plotting versus SST and TPW may conflate the processes relating these parameters in the tropics versus midlatitudes. The relationship of precipitation to TPW is well studied in the tropics (e.g., Holloway and Neelin 2009) and should mostly dominate Figs. 9 and 10 at higher SST and TPW values. The relationship between SST and precipitation is also complex and may differ in the tropics versus midlatitudes and by season (e.g., Yang and Huang 2023). Figures S6–S9 show plots similar to Figs. 9 and 10, but separately for the topics and midlatitudes that can be used for more detailed assessment of the products.
4. Conclusions
In the present work, we investigated the amount and distribution of oceanic precipitation rate from the latest GPM and GPCP products, and compared them with their previous versions, MCTG, and ERA5. We highlighted the changes through zonal plots, geographical maps, tables, and as a function of environmental variables. Here we present our main findings.
The V07 GPM Core Observatory instrument products KuPR Ku, GMI, and CMB Ku show −8.07%, +12.76%, and +9.03% change compared to their previous versions in quantifying mean oceanic precipitation rate within 65°S–65°N. GMI and CMB products show increases everywhere, but KuPR estimates do not. For PMW conical scanners (e.g., AMSR-2 and SSMIS), precipitation rate in V07 also increased everywhere compared to V06. The precipitation rate increase is also observed for PMW cross-tracking sounders (e.g., ATMS and MHS), except in midlatitudes where a slight reduction in precipitation rate is observed in both hemispheres. ATMS and MHS have similar precipitation patterns that are different from the patterns shown by AMSR-2 and SSMIS. The differences are more obvious within the 30°–50° latitude zone in both hemispheres (i.e., sounders show slightly lower precipitation rate) and near 5°N (i.e., sounders show larger precipitation rates). PMW products show a slight shift of the precipitation peak near 40°S toward higher latitudes, compared to the CMB and KuPR products. Overall, the global (90°S–90°N) ocean-mean annual precipitation rates from Version 07 of the SSMIS, MHS, ATMS, and AMSR2 are 18.01%, 14.76%, 11.90%, and 12.6% higher than their corresponding V05 mean precipitation rates. Part of the widespread increase in precipitation rates is related to addition of light precipitation rates in V07 PMW products from MiRS, where MiRS produces rain but PMW products do not.
In the tropics (25°S–25°N), the highest increase from V05 to V07 and the previous version is for MHS (19.24%) followed by ATMS (18.92%). Both products also show the highest mean precipitation rate among all other products. The substantial increase of sounders’ precipitation rate in the tropics needs further investigation, as other sensors do not show more than a 10.5% increase. CMB shows a 6.21% increase and KuPR Ku V07 shows a 9.48% decrease compared to V06. KuPR Ku is the only product that shows less precipitation than its previous version.
A closer look at the difference maps for all of the PMW (sounder and imager) products shows a narrow band of near-zero change in precipitation rate between V05 and V07 products over ocean around 55°–60°S and near the coast of Antarctic. This is related to GPROF using ERA5 precipitation over sea ice in V07, while both V05 and V07 use the CMB precipitation rates over nonfrozen regions that are much lower than ERA5 estimates in high latitudes.
Comparison of mean precipitation maps from PMW products with MCTG V07 suggests that PMW estimates are generally smaller than MCTG, but tend to be slightly larger than MCTG, KuPR, and CMB over the Pacific Ocean Intertropical Convergence Zone (centered at ∼5°N) and the northern part of the Indian Ocean. PMW sensors also show slightly higher mean precipitation rates than MCTG in the Arctic regions, where MCTG’s precipitation rate comes from CloudSat. Over the midlatitude oceans in SH (i.e., 35°–50°S), precipitation products are least consistent and display the largest spread in mean precipitation rate and location of peak precipitation. Efforts to reduce precipitation uncertainties over the midlatitude oceans in SH remain critical. In the zone covered by all products (65°S–N), MCTG V07 shows the highest mean precipitation estimate, followed by ERA5 (∼2.85% lower). GPCP V3.2 is 6.55% lower and GPM V07 products show a large range of values lower than MCTG V07, from KuPR (25.36%) to SSMIS (4.84%).
Mean monthly precipitation variation over oceans was found to be relatively small when averaged over 90°S–90°N and 65°S–65°N (e.g., ∼4%–6% of the annual mean precipitation rate), but it is larger over the tropics (e.g., 9%–12%). The variations are large when the analyses are performed separately in SH and NH for mid (25°–65°) and high (65°–90°) latitudes. In 65°–25°S, the sensors’ relative amplitude ranges from 21.51% (for SSMIS V07) to 43.81% (for KuPR V07) and are larger than that from ERA5 and GPCP V3.2. In 25°–65°N, the sensors’ relative amplitude ranges from 32.83% (for ATMS V07) to 63.84% (for GMI V07). These numbers are 60.00% and 52.47% for GPCP V3.2 and ERA5. The largest variations among the products occur in the high latitudes. In 65°–90°N, the sensors’ relative amplitude ranges from 90.48% (for SSMIS V07) to 106.61% (for MHS V07), these numbers are considerably smaller for GPCP V3.2 (67.86%) and ERA5 (65.51%). Furthermore, in this region, GPCP V3.2 and ERA5 show a different monthly precipitation distribution (e.g., in terms of timing of minimum and maximum monthly precipitation rates) than the GPM products. In 90°–65°S, the sensors’ relative amplitude ranges from 44.17% (for SSMIS V07) to 49.55% (for MHS V07), these numbers are 28.15% and 43.75% for GPCP V3.2 and ERA5 that are much smaller than their corresponding regions in NH (65°–90°N).
Precipitation products are found to be generally less consistent (larger spread) in SST and TPW bins with low or high values. At lower SST and TPW values, ERA5 shows the highest precipitation rate, followed by GPCP V3.2 and both exceeding PMW and radar products. With respect to the location of precipitation peaks and pattern versus SST and TPW, MHS and ATMS show similarities to ERA5, different from that observed by conical scanners, radar, CMB, and GPCP. While comparison of the two versions of GPM products show similar patterns, KuPR V07 tends to be lower than KuPR V06 at TPW > 24 mm and SST > 292 K and higher than V06 at lower TPWs and SSTs, where stratocumulus clouds are most frequent. As SST and TPW increase in the warming climate, it is important to make sure their relationship with precipitation is correctly captured in observational datasets to guide weather and climate models for improved predictions and projections.
The present work provides insights on areas of progress and indicates remaining discrepancies among the latest individual precipitation products. The latest estimates of ocean precipitation show some convergence of the results in terms of overall mean values and point to a general increased estimate of mean ocean precipitation. These latest individual (or dual) sensor estimates over ocean inform merged sensor analyses such as MCTG over relatively short periods (a few years), which in turn are used to inform long-term global analyses such as GPCP and shorter length, finer time-scale analyses such as CMORPH and IMERG.
Acknowledgments.
This work was carried out at The University of Arizona. Financial support is made available from NASA MEaSUREs (NNH17ZDA001N-MEASURES; Program Manager: Dr. Lucia Tsaoussi) and Precipitation Measurement Mission (NNH21ZDA001N-PMMST; Program Manager: Dr. Will McCarty) programs.
Data availability statement.
GPM and GPCP V3.2 precipitation products were downloaded from the NASA Goddard Earth Sciences Data and Information Services Center (GES DISC) (https://disc.gsfc.nasa.gov) by searching for GPCP V3.2 monthly and GPM V07 sensor products. GPCP V2.3 product was downloaded from http://eagle1.umd.edu/GPCP_CDR/Monthly_Data/. ERA5 data (precipitation, SST, and TPW) were downloaded from https://www.ecmwf.int/en/forecasts/dataset/ecmwf-reanalysis-v5.
REFERENCES
Adler, R. F., and Coauthors, 2003: The version-2 Global Precipitation Climatology Project (GPCP) monthly precipitation analysis (1979–present). J. Hydrometeor., 4, 1147–1167, https://doi.org/10.1175/1525-7541(2003)004<1147:TVGPCP>2.0.CO;2.
Adler, R. F., J.-J. Wang, G. Gu, and G. J. Huffman, 2009: A ten-year tropical rainfall climatology based on a composite of TRMM products. J. Meteor. Soc. Japan, 87A, 281–293, https://doi.org/10.2151/jmsj.87A.281.
Adler, R. F., G. Gu, and G. J. Huffman, 2012: Estimating climatological bias errors for the Global Precipitation Climatology Project (GPCP). J. Appl. Meteor. Climatol., 51, 84–99, https://doi.org/10.1175/JAMC-D-11-052.1.
Adler, R. F., G. Gu, M. Sapiano, J.-J. Wang, and G. J. Huffman, 2017: Global precipitation: Means, variations and trends during the satellite era (1979–2014). Surv. Geophys., 38, 679–699, https://doi.org/10.1007/s10712-017-9416-4.
Albergel, C., E. Dutra, S. Munier, J.-C. Calvet, J. Munoz-Sabater, P. de Rosnay, and G. Balsamo, 2018: ERA-5 and ERA-Interim driven ISBA land surface model simulations: Which one performs better? Hydrol. Earth Syst. Sci., 22, 3515–3532, https://doi.org/10.5194/hess-22-3515-2018.
Arabzadeh, A., and A. Behrangi, 2021: Investigating various products of IMERG for precipitation retrieval over surfaces with and without snow and ice cover. Remote Sens., 13, 2726, https://doi.org/10.3390/rs13142726.
Arabzadeh, A., M. R. Ehsani, B. Guan, S. Heflin, and A. Behrangi, 2020: Global intercomparison of atmospheric rivers precipitation in remote sensing and reanalysis products. J. Geophys. Res. Atmos., 125, e2020JD033021, https://doi.org/10.1029/2020JD033021.
Behrangi, A., and Y. Wen, 2017: On the spatial and temporal sampling errors of remotely sensed precipitation products. Remote Sens., 9, 1127, https://doi.org/10.3390/rs9111127.
Behrangi, A., and M. Richardson, 2018: Observed high-latitude precipitation amount and pattern and CMIP5 model projections. Remote Sens., 10, 1583, https://doi.org/10.3390/rs10101583.
Behrangi, A., and Y. Song, 2020: A new estimate for oceanic precipitation amount and distribution using complementary precipitation observations from space and comparison with GPCP. Environ. Res. Lett., 15, 124042, https://doi.org/10.1088/1748-9326/abc6d1.
Behrangi, A., M. Lebsock, S. Wong, and B. Lambrigtsen, 2012a: On the quantification of oceanic rainfall using spaceborne sensors. J. Geophys. Res., 117, 2012JD017979, https://doi.org/10.1029/2012JD017979.
Behrangi, A., T. Kubar, and B. Lambrigtsen, 2012b: Phenomenological description of tropical clouds using CloudSat cloud classification. Mon. Wea. Rev., 140, 3235–3249, https://doi.org/10.1175/MWR-D-11-00247.1.
Behrangi, A., G. Stephens, R. F. Adler, G. J. Huffman, B. Lambrigtsen, and M. Lebsock, 2014: An update on the oceanic precipitation rate and its zonal distribution in light of advanced observations from space. J. Climate, 27, 3957–3965, https://doi.org/10.1175/JCLI-D-13-00679.1.
Behrangi, A., and Coauthors, 2016: Status of high-latitude precipitation estimates from observations and reanalyses. J. Geophys. Res. Atmos., 121, 4468–4486, https://doi.org/10.1002/2015JD024546.
Behrangi, A., A. Gardner, J. T. Reager, J. B. Fisher, D. Yang, G. J. Huffman, and R. F. Adler, 2018: Using GRACE to estimate snowfall accumulation and assess gauge undercatch corrections in high latitudes. J. Climate, 31, 8689–8704, https://doi.org/10.1175/JCLI-D-18-0163.1.
Behrangi, A., A. Singh, Y. Song, and M. Panahi, 2019: Assessing gauge undercatch correction in Arctic basins in light of GRACE observations. Geophys. Res. Lett., 46, 11 358–11 366, https://doi.org/10.1029/2019GL084221.
Behrangi, A., A. S. Gardner, and D. N. Wiese, 2020: Comparative analysis of snowfall accumulation over Antarctica in light of ice discharge and gravity observations from space. Environ. Res. Lett., 15, 104010, https://doi.org/10.1088/1748-9326/ab9926.
Berg, W., T. L’Ecuyer, and C. Kummerow, 2006: Rainfall climate regimes: The relationship of regional TRMM rainfall biases to the environment. J. Appl. Meteor. Climatol., 45, 434–454, https://doi.org/10.1175/JAM2331.1.
Betts, A. K., and W. Ridgway, 1989: Climatic equilibrium of the atmospheric convective boundary layer over a tropical ocean. J. Atmos. Sci., 46, 2621–2641, https://doi.org/10.1175/1520-0469(1989)046<2621:CEOTAC>2.0.CO;2.
Bolvin, D. T., R. F. Adler, G. J. Huffman, E. J. Nelkin, and J. P. Poutiainen, 2009: Comparison of GPCP monthly and daily precipitation estimates with high-latitude gauge observations. J. Appl. Meteor. Climatol., 48, 1843–1857, https://doi.org/10.1175/2009JAMC2147.1.
Bolvin, D. T., G. J. Huffman, E. J. Nelkin, and J. Tan, 2021: Comparison of monthly IMERG precipitation estimates with PACRAIN atoll observations. J. Hydrometeor., 22, 1745–1753, https://doi.org/10.1175/JHM-D-20-0202.1.
Draper, D. W., D. A. Newell, F. J. Wentz, S. Krimchansky, and G. M. Skofronick-Jackson, 2015: The Global Precipitation Measurement (GPM) Microwave Imager (GMI): Instrument overview and early on-orbit performance. IEEE J. Sel. Top. Appl. Earth Obs. Remote Sens., 8, 3452–3462, https://doi.org/10.1109/JSTARS.2015.2403303.
Feng, X., K. Haines, and E. de Boisséson, 2018: Coupling of surface air and sea surface temperatures in the CERA-20C reanalysis. Quart. J. Roy. Meteor. Soc., 144, 195–207, https://doi.org/10.1002/qj.3194.
Hersbach, H., and Coauthors, 2020: The ERA5 global reanalysis. Quart. J. Roy. Meteor. Soc., 146, 1999–2049, https://doi.org/10.1002/qj.3803.
Hoffmann, L., and Coauthors, 2019: From ERA-Interim to ERA5: The considerable impact of ECMWF’s next-generation reanalysis on Lagrangian transport simulations. Atmos. Chem. Phys., 19, 3097–3124, https://doi.org/10.5194/acp-19-3097-2019.
Holloway, C. E., and J. D. Neelin, 2009: Moisture vertical structure, column water vapor, and tropical deep convection. J. Atmos. Sci., 66, 1665–1683, https://doi.org/10.1175/2008JAS2806.1.
Hsu, K.-l., X. Gao, S. Sorooshian, and H. V. Gupta, 1997: Precipitation estimation from remotely sensed information using artificial neural networks. J. Appl. Meteor., 36, 1176–1190, https://doi.org/10.1175/1520-0450(1997)036<1176:PEFRSI>2.0.CO;2.
Huffman, G. J., R. F. Adler, D. T. Bolvin, and G. Gu, 2009: Improving the global precipitation record: GPCP version 2.1. Geophys. Res. Lett., 36, L17808, https://doi.org/10.1029/2009GL040000.
Huffman, G. J., and Coauthors, 2020: Integrated Multi-satellitE Retrievals for the Global Precipitation Measurement (GPM) mission (IMERG). Satellite Precipitation Measurement, V. Levizzani et al., Eds., Advances in Global Change Research, Vol. 67, Springer, 343–353, https://doi.org/10.1007/978-3-030-24568-9_19.
Huffman, G. J., R. F. Adler, A. Behrangi, D. T. Bolvin, E. J. Nelkin, G. Gu, and M. R. Ehsani, 2023b: The new version 3.2 Global Precipitation Climatology Project (GPCP) monthly and daily precipitation products. J. Climate, 36, 7635–7655, https://doi.org/10.1175/JCLI-D-23-0123.1.
Iguchi, T., and R. Meneghini, 2021: GPM DPR Precipitation Profile L2A 1.5 hours 5 km V07. Goddard Earth Sciences Data and Information Services Center, accessed 3 January 2023, https://doi.org/10.5067/GPM/DPR/GPM/2A/07.
Joyce, R. J., and P. Xie, 2011: Kalman filter–based CMORPH. J. Hydrometeor., 12, 1547–1563, https://doi.org/10.1175/JHM-D-11-022.1.
Kummerow, C., 2022: GPM GMI (GPROF) Climate-based Radiometer Precipitation Profiling L2A 1.5 hours 4 km × 4 km V07. Goddard Earth Sciences Data and Information Services Center, accessed 3 January 2023, https://doi.org/10.5067/GPM/GMI/GPROFCLIM/2A/07.
Kummerow, C., W. S. Olson, and L. Giglio, 1996: A simplified scheme for obtaining precipitation and vertical hydrometeor profiles from passive microwave sensors. IEEE Trans. Geosci. Remote Sens., 34, 1213–1232, https://doi.org/10.1109/36.536538.
Lau, K.-M., and S. Shen, 1988: On the dynamics of intraseasonal oscillations and ENSO. J. Atmos. Sci., 45, 1781–1797, https://doi.org/10.1175/1520-0469(1988)045<1781:OTDOIO>2.0.CO;2.
L’Ecuyer, T. S., and Coauthors, 2015: The observed state of the energy budget in the early twenty-first century. J. Climate, 28, 8319–8346, https://doi.org/10.1175/JCLI-D-14-00556.1.
Li, Z., Y. Wen, L. Liao, D. Wolff, R. Meneghini, and T. Schuur, 2023: Joint collaboration on comparing NOAA’s ground-based weather radar and NASA–JAXA’s spaceborne radar. Bull. Amer. Meteor. Soc., 104, E1435–E1451, https://doi.org/10.1175/BAMS-D-22-0127.1.
Liu, S., C. Grassotti, Q. Liu, Y. K. Lee, R. Honeyager, Y. Zhou, and M. Fang, 2020: The NOAA Microwave Integrated Retrieval System (MiRS): Validation of precipitation from multiple polar-orbiting satellites. IEEE J. Sel. Top. Appl. Earth Obs. Remote Sens., 13, 3019–3031, https://doi.org/10.1109/JSTARS.2020.3000348.
Olson, W., 2022: GPM DPR and GMI Combined Precipitation L2B 1.5 hours 5 km V07. Goddard Earth Sciences Data and Information Services Center, accessed 3 January 2023, https://doi.org/10.5067/GPM/DPRGMI/CMB/2B/07.
Randel, D. L., C. D. Kummerow, and S. Ringerud, 2020: The Goddard Profiling (GPROF) precipitation retrieval algorithm. Satellite Precipitation Measurement, V. Levizzani et al., Eds., Advances in Global Change Research, Vol. 67, Springer, 141–152.
Rodell, M., and Coauthors, 2015: The observed state of the water cycle in the early twenty-first century. J. Climate, 28, 8289–8318, https://doi.org/10.1175/JCLI-D-14-00555.1.
Sassen, K., Z. Wang, and D. Liu, 2008: Global distribution of cirrus clouds from CloudSat/Cloud-Aerosol Lidar and Infrared Pathfinder Satellite Observations (CALIPSO) measurements. J. Geophys. Res., 113, D00A12, https://doi.org/10.1029/2008JD009972.
Sillmann, J., V. V. Kharin, X. Zhang, F. W. Zwiers, and D. Bronaugh, 2013: Climate extremes indices in the CMIP5 multimodel ensemble: Part 1. Model evaluation in the present climate. J. Geophys. Res. Atmos., 118, 1716–1733, https://doi.org/10.1002/jgrd.50203.
Skofronick-Jackson, G., and Coauthors, 2017: The Global Precipitation Measurement (GPM) mission for science and society. Bull. Amer. Meteor. Soc., 98, 1679–1695, https://doi.org/10.1175/BAMS-D-15-00306.1.
Song, Y., A. Behrangi, and E. Blanchard-Wrigglesworth, 2020: Assessment of satellite and reanalysis cold season snowfall estimates over Arctic sea ice. Geophys. Res. Lett., 47, e2020GL088970, https://doi.org/10.1029/2020GL088970.
Song, Y., P. D. Broxton, M. R. Ehsani, and A. Behrangi, 2021: Assessment of snowfall accumulation from satellite and reanalysis products using SNOTEL observations in Alaska. Remote Sens., 13, 2922, https://doi.org/10.3390/rs13152922.
Stephens, G. L., and Coauthors, 2010: Dreary state of precipitation in global models. J. Geophys. Res., 115, D24211, https://doi.org/10.1029/2010JD014532.
Stephens, G. L., and Coauthors, 2012: An update on Earth’s energy balance in light of the latest global observations. Nat. Geosci., 5, 691–696, https://doi.org/10.1038/ngeo1580.
Susskind, J., P. Piraino, L. Rokke, L. Iredell, and A. Mehta, 1997: Characteristics of the TOVS pathfinder path a dataset. Bull. Amer. Meteor. Soc., 78, 1449–1472, https://doi.org/10.1175/1520-0477(1997)078<1449:COTTPP>2.0.CO;2.
Tanelli, S., S. L. Durden, E. Im, K. S. Pak, D. G. Reinke, P. Partain, J. M. Haynes, and R. T. Marchand, 2008: CloudSat’s cloud profiling radar after two years in orbit: Performance, calibration, and processing. IEEE Trans. Geosci. Remote Sens., 46, 3560–3573, https://doi.org/10.1109/TGRS.2008.2002030.
Trenberth, K. E., L. Smith, T. Qian, A. Dai, and J. Fasullo, 2007: Estimates of the global water budget and its annual cycle using observational and model data. J. Hydrometeor., 8, 758–769, https://doi.org/10.1175/JHM600.1.
Trenberth, K. E., J. T. Fasullo, and J. Kiehl, 2009: Earth’s global energy budget. Bull. Amer. Meteor. Soc., 90, 311–324, https://doi.org/10.1175/2008BAMS2634.1.
Waliser, D. E., and N. E. Graham, 1993: Convective cloud systems and warm-pool sea surface temperatures: Coupled interactions and self-regulation. J. Geophys. Res., 98, 12 881–12 893, https://doi.org/10.1029/93JD00872.
Wang, J.-J., R. F. Adler, G. J. Huffman, and D. Bolvin, 2014: An updated TRMM composite climatology of tropical rainfall and its validation. J. Climate, 27, 273–284, https://doi.org/10.1175/JCLI-D-13-00331.1.
Xie, P., and P. A. Arkin, 1997: Global precipitation: A 17-year monthly analysis based on gauge observations, satellite estimates, and numerical model outputs. Bull. Amer. Meteor. Soc., 78, 2539–2558, https://doi.org/10.1175/1520-0477(1997)078<2539:GPAYMA>2.0.CO;2.
Yang, X., and P. Huang, 2023: Improvements in the relationship between tropical precipitation and sea surface temperature from CMIP5 to CMIP6. Climate Dyn., 60, 3319–3337, https://doi.org/10.1007/s00382-022-06519-3.
