This paper considers the place of data science in remote sensing and points out a review of relevant issues. Data science (DSci) as a discipline has been appropriately traced to the concern for data quality mostly as an input to complex decision-making situations based on examination through different approaches of very large volumes of data. The emerging massive data volumes in many fields of science, including remote sensing and other scientific and technological fields have fueled a review of the tenet of data quality assessment and analyses of the types of massive data that DSci relies on. In this paper we succinctly review these data quality concerns as a basis for our projects where a variety of massive data is considered.

Climate change is causing extreme events to become more and more frequent around the world; In particular, Mexico faces recurring floods with loss of human lives and infrastructure. The objective of this work was to identify the extent and depth of flooded areas using SAR Sentinel-1, DEM, and Random Forest and Logistic Regression algorithms. The Random Forest algorithm achieved better classification metrics over the Logistic Regression algorithm; both with a good performance to identify floods using SAR and DEM images to respond to the contingencies caused by this extreme phenomenon.

Agriculture, as a climate-dependent activity is highly susceptible to climate change. General Circulation Models (GCMs), representing physical processes in the atmosphere and land surface, are the most advanced tools currently available for simulating the response of the global climate system to increasing greenhouse gas concentrations. Maize requires favorable temperature and precipitation conditions in its different phenological stages. This crop should be recommended in areas where it expresses its maximum yield potential. The objective of this work was to generate current and future CC scenarios with representative concentration routes (RCP) of GHGs to obtain the potential yield areas of maize. The climatic scenarios obtained show climatic variation, highlighting the increase in temperature and reduction in precipitation in various regions of the study area; however, the conditions for maize crop will be more favorable in some regions, since the area with high productive potential will increase over time. This is shown by the RCP 4.5 and 8.5 models used, with a significant increase in the area of high potential from 2021 to 2080; with the medium potential, we observed a reduction of cultivated area. Low potential surfaces only show an increase with the RCP 8.5 model.

This research was carried out to identify the extent and depth of flooded areas using Sentinel-1 SAR, the Digital Elevation Model generated with LiDAR and Random Forest machine learning. Training and cross-validation was performed on a set of backscatter value samples obtained from Sentinel-1. The results indicate that out of five combinations, the Random Forest algorithm had the best performance when using the four combinations$(\text{RF}+\text{Polarization} \text{VH}+\text{VV}+$MDE) with$\text{Flm}=0.977, \text{AUC}=0.998$and$\text{Kappa}=0.955$. SAR images have potential advantages that allow rapid and efficient diagnosis of the extent of flooding caused by excess rainfall in many regions around world.

In the rainy season, Southeastern Mexico flooding is a recurrent phenomenon affecting the rural and services sector with important economic loss. The objective of this study was to detect and monitor flooding areas occurring during the rainy season using Radarsat-2 and Sentinel-1 SAR in Tabasco, Mexico. The project was carried out in 2017, a year when registered precipitation levels attained an accumulation of 2,013 mm, when September and October were the rainiest months. Cartographic information was produced concerning flooded areas, water bodies and wetlands before the waters receded. This information was compared for the same areas during the dry season. This methodology can be used to estimate the flood extents that occur frequently in this region of Mexico, especially because the phenomenon manifests itself year by year due to global warming, which is admitted, to cause greater intensity and more frequent rain events.

The State of Tabasco in Southeastern Mexico is subject to recurrent floods affecting the rural and services sector with important economic loss. The objective of this study was to contribute in the determination of flooding behavior occurring in this State during the rainy season using SAR Sentinel-1A images. The project was realized in 2017, a year when registered precipitation levels attained an accumulation of 2,013.9 mm, when September and October were the rainiest months. Cartographic information was produced concerning flooded areas, water bodies and wetlands before the waters receded. This information was compared for the same areas during the dry season. The methodology is apt to be used to monitor floods by INEGI and by state governments' personnel elsewhere in the country to respond to natural disasters and emergencies. INEGI is the official organization responsible for geographic information in Mexico.

Crop yield estimation are topics of interest in Latin-American countries, for farmers and government officers responsible of managing agricultural national policies. Besides, modern remote sensing methodologies to obtain these predictions represent important steps towards attaining the goals of precision agriculture for the 21st century. Digital data from satellite images analyzed jointly with crop modelling parameters provide information that enables crop yield estimation. The objective of this study was the estimation of yield and total volume of maize production using Spot-5 satellite images and empirical models. These models expressed a) yield (Y) as a function of LAI, and b) yield as a function of NDVI. To determine the efficiency degree of the calculated predictions at the flowering stage of the crop, yield sampling was done at the physiological maturity stage in pilot plots. Regarding yield prediction in the flowering stage, the models Y = f (LAI) reported a value of 5.96 ton.ha-1 and the model Y = f (NDVI) a value of 5.04 ton.ha-1 was obtained. These data represent 114% and 97% respectively of the true yield recorded on the field. The models are specific to the maize crop and the cultivated plots location, and that the forecasts can be acceptably accurate provided the sown areas are precisely determined.

This work was realized in the state of Guerrero, Mexico considered a high marginalization region in the country. The aim of this work was to generate the agricultural frontier and to determine the spatial distribution of maize and beans from satellite images using two classification algorithms. Of the entire state area (6,357, 781 hectares), 15.6% is occupied by agriculture. Of this, 244,585 hectares are occupied with corn and 18,034 with beans. The kappa coefficient indicates that the classification performed for Maxlike and MinDist is 91% and 89 % respectively. The results can support the design of strategies to address climate change concerns in the producing areas of corn and beans, with new varieties adaptable to current weather conditions for the benefit of farmers, particularly those located in areas of high marginality and vulnerability.

Aquaculture has an important potential in the alleviation of food shortfalls and poverty in many countries. The effective practice of aquaculture is based on several requirements, among which the location of appropriate sites and information which allows merging with activities of local populations are important. This paper reports on the location of appropriate water bodies for inland aquaculture in Mexico based on high resolution satellite images processing. The approach is illustrated for the state of Guerrero. The information is presented in thematic maps of water bodies suitable for aquaculture and the methodology employed is amenable to application in other regions.

Geospatial database items originate from the analysis of images and from the manipulation of geographic data. The corresponding datasets are described via diverse structures of metadata. GeoBase L9 is a project whose aim is to build a geospatial database to support geomatics research in agricultural and natural resource management. The first objective, to support basic browsing access to datasets, has been attained in a pilot version. This paper reports on the stage-based outlook that we have adopted towards building a semantic query facility as a medium-term objective. In particular, we examine the representation of processes applied to satellite images with respect to information items that are contained in the metadata lineage section. This query facility supports a research unit that is collectively developing and using geospatial datasets. The development of such datasets will enhance both the sharing and reuse of data by users. Further into the future, the geodatabases will able to be opened to semantic web browsers by incorporating meaning in the metadata.

The objective of the project reported in this paper was to study land use change over a period of 23 years using Landsat satellite images using a methodology based on territorial transformation. An analysis of land use change was realized from 1997 to 2000 in the municipality of Texcoco in Central Mexico. The analysis unit was determined by landscape, since modifications that have taken place over the study period are due to the dynamic effects of territorial occupation. The morphological situation was interpreted as it relates to structure and functional level in order to explain the territorial changes observable in this municipality. The land use change in the study area has intensified due to new transportation networks, urban growth and the current plan to extend the international Mexico City airport toward this municipality.

Compact agricultural areas in Mexico have been identified, which are monitored as to their behavior concerning production and rural productivity in a network of Agrotech Observatories (AOTs). An AOT is a compact agricultural area representative of agro-ecological, technological and social conditions in the country. A multidisciplinary team of scientists and researchers analyze and define the best production options for the different types of producers in these areas. To optimize production and agricultural productivity in compact areas, a multidisciplinary and holistic approach with four lines of activity (agro-ecological, technological, economic, and social), and ten actions are used. One of them is oriented towards determination of the land use/cover over sixteen compact agricultural areas in Mexico. Currently, it is important to have updated and accurate information to support actions and programs of federal, state and local government for farmers, particularly in compact areas with high agricultural production potential.

In central Mexico, specifically in the State of Mexico maize is cultivated under different technological regimes ranging from traditional rain water dependency and native seeds producing yields of under 1.0 ton/ha, up to irrigation and improved seeds regimes with yields above 12.0 ton/ha. The average state yield harvested area for this crop in the past 10 years has been 549, 000 ha with an average yield of 3.22 tons/ha and an overall average of 1.8 million tons of grain. To obtain the cultivated area, SPOT panchromatic and multispectral satellite images were processed over the growth and development stages of the plants. In the cultivated areas sample yield data were collected, geo-referencing the collection sites with geographic information management products. These data were spatially represented via interpolation. The final products were yield maps at different cartographic scales.

There is a need in Mexico for accurate and up to date forest inventories due to concerns related with sustainable development programs. Forest inventories in the past have been incomplete and are not useful at regional levels. The national forest inventory is scheduled to be updated soon -optical remote sensing techniques and traditional field surveying are envisaged. There are no operational applications of radar remote sensing for forest management within government agencies or academic institutions in the country. Forest land cover is dynamical due to urban area growth, illegal logging and forest clearing for agricultural purposes in many regions. A previous land cover project combined Landsat-ETM and RADARSAT-1 imagery in Central Mexico, where forest areas are frequently foggy. A current project involves testing the potential benefits of combining polarimetric radar and optical data for forest applications. RADARSAT-2 imagery will be used as part of the Science and Operational Applications Research Program. The project aims to evaluate combined optical/radar approaches to improve forest inventory at regional scales. As a first step the total forested area is determined from optical SPOT 5 images. We show preliminary results from the optical data which are being validated on the field. As RADARSAT-2 imagery become available, polarization signatures for forest parameters will be obtained. Further work will evaluate the complementarities with optical signatures in determining forest species.

In agriculture, soil and crop conditions change from day to day and throughout the growing season. Agricultural targets also vary spatially with differences observed from field to field, as well as within individual fields. The heterogeneity of corn-growing conditions in Mexico makes accurate data for crop type, crop condition and crop yield prediction difficult to obtain. Yield predictions are needed by the federal government to estimate, ahead of harvest time, the amount of corn required to be imported in order to meet the expected domestic shortfall. In this project a methodology for the estimation of corn yield ahead of harvest time is developed which uses radar and optical remote sensing and which specifically considers the corn-growing situation in Central Mexico. Radar based crop type classification requires data sets with multiple polarizations. Recent research to assess relative classification accuracies of multi-polarized combinations for target crops using airborne data has been reported. In addition to identifying crop type and variety, identifying crop growth stage is valuable. Crop condition, loosely defined as the vigor or health of a crop in a particular growth stage, is related to crop productivity and yield; however, the relationship is complex. Main crop condition indicators include biomass, height, leaf area and contents of plant water, chlorophyll and nitrogen. Crop-type and crop-condition mapping are among the applications that are expected to benefit the most from the technical enhancements embodied by RADARSAT-2. The potential of RADARSAT-1 data for these applications has been rated as "limited", whereas for RADARSAT-2 data this potential is anticipated to be "strong". The Science and Operational Applications Research for RADARSAT-2 Program (SOAR) is promoting the evaluation of Synthetic Aperture Radar (SAR) capabilities by providing images to selected research projects which include the present one.objectives of this project are: a) use RADARSAT-2 data and optical data to determine cultivated areas and monitor crop condition for obtaining better estimations of crop yield; b) obtain polarization signatures from RADARSAT-2 data for corn and relate these to Leaf Area Index and photosynthetic active radiation (PAR) crop parameters and vegetation indexes, to establish indicators of crop condition and produce estimates for crop yield; c) use field data collected for three key corn crop growth stages over 300 pilot plots during 2001-2006, and increase the number of plots to build a database to support accuracy studies using RADARSAT-2 data.The expected benefits of this project are: to obtain knowledge about crop type, crop condition and crop yield with better accuracy than with current methodologies; to support national corn farmers associations; to design agriculture related activities within State agriculture plans; to support the corn product industry and aid government decision making. Relevant results and economical impact will imply operational usage of RADARSAT- 2 data in the agricultural sector in Mexico.