High-resolution air temperature dataset for Peru

A gridded national product that merges weather station, reanalysis, topographic and remote-sensing data.
High-resolution air temperature dataset for Peru
Like

Air temperature is a key climate parameter with a wide range of applications across disciplines. From guiding our clothing choices to shaping agricultural practices, and contributing to our understanding of climate change, air temperature is a crucial variable with broad significance at the nexus of short-term decision-making and long-term environmental planning. Usually, air temperature is obtained from weather stations with high accuracy and temporal resolution. However, detailed region-wide availability of temperature data is limited due to unequal distribution and limited time series of weather stations. This is particularly true for high-mountain regions with complex topography and limited resources for climate monitoring such as Peru. In these regions, gridded global- or continental-scale datasets are frequently used but limitations related to the complex topography, spatial resolution and the amount of assimilated data reduce their reliability. To overcome these shortcomings, the National Service of Meteorology and Hydrology (SENAMHI) has developed a gridded air temperature product as part of the Peruvian Interpolated data of the Climatological and Hydrological Observations of SENAMHI (PISCO), denominated PISCOt v1.1. Since its release, PISCOt has been applied in diverse fields of research and practice [1, 2]. Due to the increasing availability of weather data and the need for detailed spatial information for high-resolution distributed modeling approaches, we developed the new PISCOt v1.2 product. This novel dataset has substantially improved compared to its earlier version in several key aspects: more assimilated time series, better consistency of station data pre-processing, use of updated freely available auxiliary predictors, higher spatial resolution (0.01° compared to 0.1° of PISCOt v1.1), a tidier and revised calculation sequence, and improved version control. Therefore, the development of PISCOt v1.2 is more consistent, traceable, and reproducible compared to other previously established air temperature gridded products in Peru. This endeavor represents an important baseline for high-resolution climate modeling including applications related to climate change, hydrology and ecology, among others.

How was PISCOt v1.2 developed?

In the study published last week in Scientific Data (https://www.nature.com/articles/s41597-023-02777-w) by Nature Research, the development of the high-resolution air temperature dataset PISCOt v1.2 for entire Peru (1981-2020) was outlined following a four-step process: i) quality control, ii) gap-filing, iii) homogenization and iv) spatial interpolation. In the first three steps the focus was on creating a corrected and consistent air temperature database. This has been a challenging procedure given the huge amount of non-quality-controlled, missing and inhomogeneous weather station data. Once the complete database was established, gridded air temperature for entire Peru was estimated using spatial interpolation based on kriging techniques with topographic and remote sensing (MODIS Land Surface Temperature) predictors. 

To guarantee the highest consistency of PISCOt v1.2, it was necessary to consider several critical points. First, enhanced visualizations techniques to improve data quality control of all air temperature time series were used. Second, to avoid artificial inhomogeneities in the final gridded dataset, a gap-filling algorithm was developed to reconstruct air temperature time series back to 1981 by employing nearby stations and the ERA5-Land reanalysis dataset in an iterative way. Third, as MODIS Land Surface Temperature was only available since the 2000s, the  climatologically aided interpolation approach was used where the spatial interpolation was divided into the mean monthly normal and anomalies, and then aggregated to obtain the final product. Finally, as Peru presents high steep climate gradients non-stationary and stationary spatial models were applied in combination of the above-mentioned covariables, leading to a proper account of the local spatiotemporal variability in the estimation of air temperature for the whole country.

The new PISCOt v1.2 was evaluated in multiple ways in order to check its potential benefits. From a methodological viewpoint, the gap-filling process and the spatial model were validated and cross-validated at monthly normal (mean) and daily scale. In addition, the usefulness of this new dataset was tested through two applications: one associated with spatial-temporal variability of air temperature plus other gridded air temperature products (ERA5-Land, TerraClimate CHIRTS, among others; Fig 1), and a second one related to coastal fog effects on air temperature. Both applications proved to be satisfactory for both the gap-filling and spatial models. The use of MODIS Land Surface Temperature was crucial for achieving higher predictivity of minimum air temperature at monthly normal and daily scale.

Fig 1. Annual rate of change (1983-2013) of temperature indices (mean Tmax (MTmax), mean Tmin (MTmin), and frost days (FD)) for PISCOt v1.2 and gridded products (PISCOt v1.1, VS2018, TerraClimate, CHIRTS, and ERA5-Land) over southern Andes of Peru. Black lines represent three main regions: Pacific Coast, Andes, and Amazon; Lake Titicaca is shown as a light blue filled area.

Where is PISCOt v1.2 stored?

The new air temperature dataset at high-resolution (0.01°) from 1981 to 2020 for the entire Peru is publicly available as a Figshare collection (https://doi.org/10.6084/m9.figshare.c.5959863). Moreover, this gridded product is also available at a coarser spatial resolution (at 0.05° and 0.10°) to comply with other specific data and modeling needs.

The PISCOt v1.2 database is a new valuable dataset for different applications in Peru as it allows for high-resolution analyses linked to e.g. climate change, health, hydrology, ecosystem assessments, and other fields for research and practitioners. PISCOt v1.2 supports the generation of new findings urgently required for more robust local decision-making in the scientific and political communities, especially in a context of data scarcity and high uncertainties that can commonly be found in Peru and the region.

References

[1] Delahoy, M. J., Cárcamo, C., Huerta, A., Lavado, W., Escajadillo, Y., Ordoñez, L., ... & Levy, K. (2021). Meteorological factors and childhood diarrhea in Peru, 2005–2015: a time series analysis of historic associations, with implications for climate change. Environmental Health, 20(1), 1-10. https://doi.org/10.1186/s12940-021-00703-4 

[2] Motschmann, A., Teutsch, C., Huggel, C., Seidel, J., Leon, C. D., Munoz, R., ... & Weimer-Jehle, W. (2022). Current and future water balance for coupled human-natural systems–Insights from a glacierized catchment in Peru. Journal of Hydrology: Regional Studies, 41, 101063. https://doi.org/10.1016/j.ejrh.2022.101063

Please sign in or register for FREE

If you are a registered user on Research Communities by Springer Nature, please sign in

Subscribe to the Topic

Climate Sciences
Physical Sciences > Earth and Environmental Sciences > Earth Sciences > Climate Sciences
Climate and Earth System Modelling
Mathematics and Computing > Mathematics > Applications of Mathematics > Mathematics of Planet Earth > Climate and Earth System Modelling
Earth Sciences
Physical Sciences > Earth and Environmental Sciences > Earth Sciences
Statistics in Engineering, Physics, Computer Science, Chemistry and Earth Sciences
Mathematics and Computing > Statistics > Applied Statistics > Statistics in Engineering, Physics, Computer Science, Chemistry and Earth Sciences

Related Collections

With collections, you can get published faster and increase your visibility.

Meteorology and hydroclimate observations and models

This Collection presents a series of articles describing hydroclimate datasets, including data sourced from remote sensing, primary measurements or theoretical models. Datasets are presented without analyses in order to support policy development and further research, with Data Descriptors providing full details of data sources, modelling, and any associated code.

Publishing Model: Open Access

Deadline: Dec 15, 2023

Medical imaging data for digital diagnostics

This Collection presents a series of articles describing annotated datasets of medical images and video. All medical specialities are considered and data can be derived from study participants, tissue samples, electronic health records (EHRs) or other sources.

Publishing Model: Open Access

Deadline: Dec 20, 2023